SPATIOTEMPORAL CONTROL OF THE SALT STRESS INDUCED TRANSCRIPTIONAL RESPONSE IN ARABIDOPSIS

Transcription

1 SPATIOTEMPORAL CONTROL OF THE SALT STRESS INDUCED TRANSCRIPTIONAL RESPONSE IN ARABIDOPSIS RUI WU NATIONAL UNIVERSITY OF SINGAPORE 2013

2 SPATIOTEMPORAL CONTROL OF THE SALT STRESS INDUCED TRANSCRIPTIONAL RESPONSE IN ARABIDOPSIS RUI WU A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY DEPARTMENT OF BIOLOGICAL SCIENCES NATIONAL UNIVERSITY OF SINGAPORE 2013

3 DECLARATION I hereby declare that the thesis is my original work and it has been written by me in its entirety. I have duly acknowledged all the sources of information which have been used in the thesis. This thesis has also not been submitted for any degree in any university previously. Rui Wu August 20th, 2013

4 ACKNOWLEGEMENT The first big thank you I want to send to my supervisor, Dr. Jose R. Dinneny. Thanks for providing the good research environment and challenging ideas during my PhD study; positive attitude and kind encouragement when I encountered depression and frustration from the projects and life; as well as his kind and open support for my decisions on both work and life, just like a friend. I became more independent as a scientific thinker and problem solver, and closer to a real researcher. Thanks to the companionship of my kind and supportive lab mates, I did not feel alone abroad doing this challenging thing. Whenever I needed help, they gave me hands without hesitation. Thanks to Lina for her generous help and guidance when I was first in the lab and her companion for the entire 4 years. Thanks to Jeffrey, Chonghan, Xie Fei, Pooja, Geng Yu, Shahram, MC, Neil, Bao Yun, Ruben and Jose for the valuable discussions and advice for my projects. Thank Penny, Han-qi and Cliff for their work assisting my study. Thanks to all intern students and undergraduate students doing the final year projects for the work they have done. I would like to thank the Department of Biological Sciences in NUS for providing the precious opportunity for me to pursue my PhD degree; it is really a great university that gives access to advanced research, excellent people and valuable resources. And I would also like to thank Temasek Life Sciences Laboratory and Carnegie Institution of Plant Sciences for providing the facilities and platform for me to do my study and communicate with excellent people. Thank all my friends in Singapore and the US for making my life abroad like at home. Thanks to our collaborators, Dr. Jose Pruneda-Paz and Dr. Steve Kay, for their efforts on transcription factor screening. And thanks to Dr. Hao Yu lab (Temasek Lifesciences I

5 Laboratory) for providing the vectors for GUS reporter and seeds of RGA::GFP:GRA. Thanks to Dr. Joseph Horecka in Prof. Ronald Davis lab (School of Medicine Department of Biochemistry, Stanford University) for his generous advice and reagents for yeast transformation and colony PCR. At last, I want to give the biggest Thank you to my family. Thank you for giving me continuous love and support. This will be the most precious gift in my life. July 27, 2013 Rui Wu II

6 TABLE OF CONTENTS ACKNOWLEGEMENT...I TABLE OF CONTENTS... III SUMMARY... VII LIST OF TABLES... IX LIST OF FIGURES... X CHAPTER 1 LITERATURE REVIEW HIGH SALINITY STRESS IN PLANTS High salinity affects different developmental events of plants Evolutionary variations of plant adaption to high salinity stress Halophytes Glycophytes Secondary physiological responses involved in high salinity stress Hyper-osmotic stress Dehydration (drought stress) Ion disequilibrium Oxidative stress Hormone involvement in salt response Abscisic acid (ABA) Ethylene Gibberellic acid (GA3) Brassinosteroids III

7 Cytokinin Auxin Studies of high salinity stress in Arabidopsis Arabidopsis is a model plant in salt stress studies Root a multicellular organ directly responsive to salt stress TRANSCRIPTIONAL REGULATION AND TRANSCRIPTIONAL NETWORK Transcriptional regulation is an indispensable process involved in developmental process and environmental stimuli response Mechanisms of transcriptional regulation Approaches to generate a transcriptional network OBJECTIVES AND SIGNIFICANCE OF THIS STUDY CHAPTER 2 MATERIALS AND METHODS PLANT MATERIALS PLANT GROWTH CONDITIONS AND STRESS TREATMENT GENERATION OF TRANSGENIC LINES Sequences design of synthetic promoters Constructs Agrobacterium mediated plant transformation YEAST ONE HYBRID SCREEN Constructs generation Yeast transformation Yeast one hybrid screening BIOINFORMATICS DATA ANALYSIS LIVE IMAGING CONFOCAL MICROSCOPIC ANALYSIS IV

8 2.8 GUS STAINING LUC ANALYSIS GENE EXPRESSION GENETIC ANALYSIS CHAPTER 3 RESULTS AND DISCUSSIONS I ABSTRACT INTRODUCTION RESULTS A global spatiotemporal transcriptional map of the salt stress response in Arabidopsis root Different strategies were used to adapt salt stress at early and late stages A cluster-comparison method identifies targets mediating hormone signaling in salt stress response Spatiotemporal understanding of hormone biosynthesis and signaling pathway - --ABA as an example ABA signaling mediated transcriptional response to salt stress showed tissue specificities Dynamic involvement of GA signaling during salt stress response DISCUSSIONS CHAPTER 4 RESULTS AND DISCUSSIONS II ABSTRACT INTRODUCTION RESULTS V

9 4.3.1 Schematic description of the pipeline for setting up the transcriptional network Identification of the salt responsive cis-regulatory elements based on the spatiotemporal transcriptional map of Arabidopsis roots Synthetic promoters harboring CREs confer the ability to drive specific expression patterns under normal or stress conditions Synthetic promoters containing CREs confer the ability to respond to salt stress in a dynamic manner Synthetic promoter strategy for screening using the TF library DISCUSSION Synthetic promoters drive tissue-specific and salt responsive patterns ABRE s expression pattern indicates the location of ABA signaling in root development and environmental response Combinatorial properties of regulatory elements necessary for environmental stress response CHAPTER 5 CONCLUSIONS REFERENCES APPENDIX CURRICULUM VITAE VI

10 SUMMARY For all living things, the ability to respond to environmental stress is an essential property. Various environmental stimuli can be processed by organisms, resulting in different kinds of responses, such as morphological and physiological changes as well as actual behaviors. With these responses, organisms can acclimate effectively for survival. Different from animals that can escape from a poor environment, the only strategy plants can use is to acclimate. However, an organism is just like a black box, because how the input environmental signals are processed is not clear, but what is known is that intricate signal transduction and transcription networks must be involved. My study is focused on how different signaling pathways are integrated spatiotemporally under high salinity stress and how transcriptional regulation occurs. Firstly, I did an analysis on a previously generated spatiotemporal transcriptional map of salt stress in Arabidopsis roots, covering 4 core cell types and 6 time points for salt treatment. Compared with the previous study showing tissue-specific responses at 1 hour to high salinity, this map provided higher temporal resolution, giving a more dynamic view of how different cell types respond to salt stress at different time periods of salt treatment. Based on this spatiotemporal map, the transcriptional changes of key components in hormone biosynthesis and signaling were identified, suggesting that these hormones function in specific cell types and at particular stages during acclimation to high salinity. A bioinformatics method was also developed to systematically de-convolve the hormone crosstalk network with salt stress, identifying some salt stress response submodules controlled by hormone signaling. A good portion of these modules were validated using high throughput q-rt PCR. The dynamic transcriptional regulation and homeostasis mediated by hormone signaling is well correlated to the dynamic root growth illustrated by my colleague. VII

11 Second, complex transcriptional networks composed of cis-regulatory elements (CREs) and their corresponding transcription factors (TFs) allow us to understand how higher plants are normally developed and transcriptionally respond to environmental stimuli. Although, in the past, numerous putative CREs were computationally predicted, only a few were experimentally verified with their biological functions. Here, I developed an efficient pipeline to study the biological functions of cis-regulatory elements which are good starting points for the generation of a CRE centered transcriptional network involved in the salt stress response in the Arabidopsis roots. The pipeline includes: bioinformatics search and functional validation of CREs, high-throughput screening of TFs binding the CREs via yeast one hybrid and the functional validation of the TFs, as well as generation of a transcriptional network. Using this pipeline, I have validated the regulatory functions of seven CREs, including ABRE (ABA response element), which is known to be involved in salt and drought stresses, and two other previously unknown elements. The strategy I used is useful and efficient in studying the biological functions of CREs and provides a good starting point for promoter engineering in the future. In addition, the parameters for this approach were tested systematically to get an optimal method for future use. VIII

12 LIST OF TABLES TABLE 1. THE MULTIMERIZED UNIT SEQUENCES FOR SYNTHETIC PROMOTERS TABLE 2. PRIMER SEQUENCES USED IN CLONING, SEQUENCING AND COLONY PCR TABLE 3. ACCESSION NUMBERS OF ANALYZED GENES AND PRIMERS SEQUENCES USED DURING THE REAL-TIME QUANTITATIVE PCR ANALYSIS TABLE 4. TRANSCRIPTION FACTORS SHOWING OVERLAP BETWEEN THE TWO VERSIONS OF SYNTHETIC PROMOTERS FROM Y1H SCREENING IX

13 LIST OF FIGURES FIGURE 1. SCHEMATIC LONGITUDINAL AND CROSS SECTION OF ARABIDOPSIS ROOT TIP (ADAPTED AND MODIFIED FROM DINNENY ET AL., 2008) FIGURE 2. GENERATION OF SPATIOTEMPORAL TRANSCRIPTIONAL MAP FIGURE 3. EXPRESSION OF DEVELOPMENTAL GENES IN THE SPATIOTEMPORAL MAP UNDER SALT STRESS FIGURE 4. PRINCIPAL COMPONENT ANALYSIS OF THE DIFFERENT SAMPLE TYPES COMPOSING THE SPATIO-TEMPORAL MAP OF THE SALT RESPONSE FIGURE 5. NUMBER OF GENES THAT SHOWED DIFFERENTIAL EXPRESSION IN EACH CELL LAYER AT DIFFERENT TIME POINTS AFTER SALT TREATMENT FIGURE 6. SPATIOTEMPORAL EXPRESSION PATTERNS OBSERVED DURING THE SALT RESPONSE FIGURE 7. BIOLOGICAL PROCESSES REGULATED IN SPATIOTEMPORAL SALT STRESS RESPONSE FIGURE 8. BIOLOGICAL PROCESSES INVOLVED IN EARLY AND LATE STAGES OF SALT STRESS RESPONSES IN DIFFERENT CELL TYPES FIGURE 9. ANALYSIS OF THE HORMONE SECONDARY SIGNALING NETWORK REGULATING SALT-DEPENDENT TRANSCRIPTIONAL PROGRAMS FIGURE 10. ABA PLAYS ROLES IN EARLY STAGE OF SALT STRESS RESPONSE FIGURE 11. POTENTIAL CROSSTALK BETWEEN ABA AND CYTOKININ FOR THE REGULATION OF THE GENE EXPRESSION AT EARLY STAGE FIGURE 12. ABA BIOSYNTHESIS IS REGULATED IN EARLY STAGE OF SALT STRESS RESPONSE FIGURE 13. CELL LAYER SPECIFIC ABA SIGNALING REGULATES SPATIALLY LOCALIZED TRANSCRIPTIONAL CHANGES X

14 FIGURE 14. GA SIGNALING WAS DYNAMICALLY REGULATED DURING SALT STRESS FIGURE 15. SCHEMATIC CHART SHOWING THE WORKFLOW FOR THE SYNTHETIC PROMOTER APPROACH FIGURE 16. THE IDENTIFICATION OF KNOWN ELEMENTS ENRICHED WITH THE 25 SALT CLUSTERS USING THE METHOD OF ATHENA FIGURE 17. THE IDENTIFICATION OF KNOWN ELEMENTS ENRICHED WITH THE 25 SALT CLUSTERS USING THE METHOD OF FIRE FIGURE 18. THE IDENTIFICATION OF THE TISSUE-SPECIFIC ELEMENTS USING THE HIGH RESOLUTION SPATIAL MAP FIGURE 19. SYNTHETIC PROMOTERS HAVE THE ABILITY OF DRIVING SPECIFIC EXPRESSION PATTERNS FIGURE 20. SALT RESPONSIVE ELEMENTS FOR THE FURTHER ANALYSIS FIGURE 21. THE EXPRESSION OF ABRE SYNTHETIC PROMOTER FIGURE 22. SYNTHETIC PROMOTER CONFERS THE ABILITY OF RESPONDING TO ENVIRONMENTAL STRESSES FIGURE 23. QUANTIFICATION OF GUS REPORTER DRIVEN BY ABRE SYNTHETIC PROMOTERS FIGURE 24. EXPERIMENTAL TEST OF ALTERNATIVE MINIMAL PROMOTER INSTEAD OF 35S MINIMAL PROMOTER FIGURE 25. TEST OF THE EFFECT OF FLANKING SEQUENCES AND REPEAT NUMBER FOR THE SYNTHETIC PROMOTER FIGURE 26. EXPRESSION PATTERNS OF THE 2 UNKNOWN SALT RESPONSIVE CIS-REGULATORY ELEMENTS FIGURE 27. DRE SYNTHETIC PROMOTERS SHOWED VARIABLE EXPRESSIONS UNDER NORMAL CONDITION XI

15 FIGURE 28. L1 BOX SHOWED DIFFERENT EXPRESSION PATTERNS BETWEEN THE TWO DIFFERENT VERSIONS OF SYNTHETIC PROMOTERS FIGURE 29. EXPERIMENTAL DESIGN FOR THE DYNAMIC RESPONSE OF SYNTHETIC PROMOTERS UNDER SALT STRESS FIGURE 30. EFP SHOWING THE SPATIOTEMPORAL EXPRESSION PATTERN OF UBQ10 UNDER SALT STRESS REPONSE FIGURE 31. ABRE SYNTHETIC PROMOTERS RESPOND TO SALT STRESS DYNAMICALLY FIGURE 32. DYNAMIC ANALYSIS OF SALT STRESS RESPONSE OF THE KNOWN ELEMENTS FIGURE 33. DYNAMIC ANALYSIS OF SALT STRESS RESPONSE OF THE UNKNOWN ELEMENTS FIGURE 34. EXPERIMENTAL TEST OF DIFFERENT VERSIONS OF ABRE SYNTHETIC PROMOTERS FOR TF SCREENING USING Y1H FIGURE 35. PRINCIPLE COMPONENT ANALYSIS SHOWED THE EFFECT OF FLANKING SEQUENCE BASED ON THE TF BINDING AFFINITY FIGURE 36. TF BINDING AFFINITY COMPARISON BETWEEN DIFFERENT TEST VERSIONS OF ABRE FIGURE 37. IN VIVO VALIDATION OF THE INTERACTION BETWEEN ABRE SYNTHETIC PROMOTER AND THE TRANSCRIPTION FACTORS OBTAINED FROM YEAST ONE HYBRID. 125 FIGURE 38. PROTEIN STABILIZATION OF BZIP FAMILY TRANSCRIPTION FACTORS UNDER SALT STRESS FIGURE 39. PROTEIN STABILIZATION OF C2H2 FAMILY TRANSCRIPTION FACTOR, AZF3, UNDER SALT STRESS XII

16 LIST OF ABBREVIATIONS AND SYMBOLS Unites g M hr min μg μl μm μm ml mm nm mm nm kv gram molar Hour minute microgram microlitre micrometer micromolar milliliter millimolar nanometer millimeter nanomolar kilo volt C degree celsius bp kb rpm w/v base pairs kilo base-pairs revolution per minute weight per volume Chemicals and reagents ABA Abscisic Acid XIII

17 ACC IAA GA EtOH dh 2 O MES KOH NaCl PAC KCl MgSO 4 1-aminocyclopropane-1-carboxylic-acid 3-Indoleacitic acid Gibberellic Acid Ethanol Distilled water 2-(N-Morpholino)ethanesulfonic acid Potassium hydroxide Sodium chloride Paclobutrazol Potassium chloride Magnesium sulfate XIV

18 Chapter 1 Literature review 1

19 1.1 High salinity stress in plants Two decades ago, it was estimated that at least 20% of the world's arable land and more than 40% of irrigated land were affected by high salinity (Rhoades and Loveday, 1990). Nowadays, this problem is much more serious. High salinity has become the most common agricultural contaminant in the world, affecting the yield and quality of many crops, to the detriment of an over-increased population. Many processes are affected by high salt concentration, such as seed germination, seedling and vegetative growth, and flowering etc. (Sun and Hauser, 2001; Xiong et al., 2002; Macler and MacElroy 1989). Plants are classified as glycophytes and halophytes based on their capacity to grow on high concentration salt medium (Flowers et al., 1977). Most plants, including the majority of crops, are glycophytes and cannot tolerate salt-stress, while halophytes are native flora of saline environment (Flowers et al., 1986). High salt concentrations do harm to glycophytes mainly by causing cellular and physiological changes that produce secondary effects, such as hyper-osmotic stress, dehydration, oxidation, and ion disequilibrium, as well as cytotoxicity (Hasegawa et al., 2000; Zhu, 2001), which will be further discussed. However, when facing environmental stresses such as salt stress, plants can develop mechanisms, such as different hormone signaling and their downstream signals, adapting or 'micro-avoiding' the threats. Due to the complexity of an organism, the mechanisms through which plants achieve this purpose are complicated. In these processes, cells with different identities perform differently in response to stress, due both to the spatial positions of the cells and to the biological functions of the cells. So, how a multicellular organism dynamically interprets environmental stresses will provide a better understanding of the mechanisms for salt response, adaption and tolerance High salinity affects different developmental events of plants 2

20 Sodium chloride is the major contaminant for salt stress. Its toxicity for plants mainly lies in the following aspects. Firstly, high salt concentration decreases the osmotic potential of the soil solution and reduces the water potential of cells, thus leading to water stress in plants. Secondly, ionic toxicity is caused, since excessive Na + cannot be readily sequestered into vacuoles and thus changes the ratio of Na + /K +, leading to a nutrient deficiency in K +. The direct consequence for plants is the disruption of many developmental processes. Under mild salt stress, plant cells dehydrate and shrink due to the lower water potential and regain their original volume hours later after acclimation. But cell elongation and cell division in this process are still reduced, leading to lower rates of leaf and root growth (Munns, 2002). A recent study showed that the growth rate of lateral roots is also affected by salt stress dynamically, and a quiescent phase happens very quickly after salt stress, as observed and quantified by live imaging (Duan et al., 2013). Other than this quick effect in reduction of growth, long-term reduced growth and even leaf death occurs. This is the result of salt accumulation in leaves, which causes the death of leaves and reduction of the total photosynthetic leaf area (Munns, 2002). This long-term effect cannot be recovered from. In addition, salt stress also affects other developmental processes. Several studies have indicated that high salinity not only delays germination but also reduces the percentage of germinated seeds (Carter et al., 2005; Mauromicale and Licandro, 2002). Also, the reproductive process is affected by salt stress. For example, a study on rice indicated that salinity results in delayed flowering and reduces the number of productive tillers, fertile florets per panicle and individual grain weight (Khatun et al., 1995; Lutts et al., 1995). From this, we know that high salinity affects the yield of crops to a great extent, so studies on mechanisms of salt stress response and tolerance are necessary. In the following, an introduction to these studies will be made. 3

21 1.1.2 Evolutionary variations of plant adaption to high salinity stress As a result of different evolutionary strategies, plants can be categorized into two groups, glycophytes and halophytes. Halophytes can grow on salt concentrations as high as over 400 mm, which is about 10 times that tolerated by glycophytes (Flowers et al., 1977). The differences between halophytes and glycophytes with respect to salt tolerance mechanism were summarized (Parida and Das, 2005) Halophytes A halophyte is a plant that is adapted to grow in soil with high salinity, such as in saline semi-deserts, mangrove swamps, marshes and sloughs, and seashores. The mechanisms for salt tolerance in halophytes have been studied, and structures called salt glands were found to be important for halophytes to secrete excess salt ions, which is salt contaminant causing toxicity (Labidi et al., 2010). In addition, study of amino acid content in halophytes and glycophytes suggests that osmolytes can be another important factor for salt tolerance. For example, proline accumulates in halophytes at a much higher level when induced by salt treatment (Stewart and Lee, 1974). Although many studies provide information about factors that contribute to salt tolerance, the underlying mechanisms in halophytes are still largely unknown (Flowers and Colmer, 2008). The development of high-throughput DNA sequencing technologies allows us to understand the evolutionary patterns that are at the basis of halophytic adaptations to extreme environments and the mechanisms for salt tolerance. For example, the genome sequences for Thellungiella. salsuginea and Thellungiella. parvula has been available (Dassanayake et al., 2011). Although they are still in the form of chromosome models, the analysis of the sequences reveals some specific properties different from A. thaliana, like the movement of 4

22 centromeric regions and difference in TE (transposable element) proliferation and regulatory sequences upstream of coding regions (Dassanayake et al., 2011) Glycophytes Unlike halophytes, glycophytes are more sensitive to salt stress. Although they might not have as strong adaptation mechanisms to salt as halophytes, their sensitivity to salt allows us to explore the changes inside the cell environment in order to further investigate salt tolerance mechanisms. For instance, it was found that in glycophytes, the toxicity effect mainly comes from the accumulation of Na + in leaves. The built-up ions in the cytoplasm of leaf cells will inhibit enzyme activity and lead to senescence (Munns and Passioura, 1984; Flowers and Yeo, 1986). This process is regulated by Na + transporters, including the initial entry into the roots through some non-selective cation channels or high affinity K + transporters (Shabala et al., 2007), and the transfer from root to shoot, including a Na + transporter, HKT1 (Davenport et al., 2007). In the following introduction, I will review studies in glycophytes Secondary physiological responses involved in high salinity stress Salt stress response is a very complicated process involving many different secondary stresses, as plants have evolved complex signaling pathways in response to various stimuli, such as salt, osmosis, drought, oxidative stress. Previous studies have suggested that cell signaling pathways can be shared by these different stress events, with the same stress perception sensors, the same secondary signal, like Ca 2+, and the same regulatory elements, etc. (Chinnusamy et al., 2004). Also, cross-talk between theses pathways may 5

23 reveal a common stress induced signaling pathway and supply a basis for the mechanisms of environmental stresses Hyper-osmotic stress Hyper-osmotic stress is the most immediate consequence of high salinity. When the root encounters a saline solution, the chemical potential establishes a water potential imbalance between the apoplast and symplast, and this imbalance leads to a decrease in turgor pressure, which causes a growth reduction if severe (Bohnert et al., 1995). To relieve osmotic stress, plants have developed several mechanisms, such as the Ca 2+ signaling mediated SOS pathway to exclude Na + ions out of cells, compatible osmolytes and osmoprotectants to increase the turgor pressure, and Na + vacuolar compartmentalization, decreasing cytosolic sodium ions (Yokoi et al., 2002). Accumulation of osmolytes and osmoprotectants can serve as a long-term strategy against hyper-osmosis because these compounds can accumulate to high levels without disturbing intracellular biochemistry. The compounds include simple sugars (e.g. fructose and glucose), sugar alcohols (e.g. glycerol) and complex sugars (e.g. fructans). Some amino acid derivatives, like proline, glycine betaine, polyols and proline betaine, also meet this need. For example tobacco plants transformed with bacterial glycine betaine biosynthesis genes showed accumulated glycine betaine and higher salt tolerance (Holmstrom et al., 2000). Another example suggested that the expression of bacterial choline oxidase gene CodA in Arabidopsis caused glycine betaine accumulation and increased tolerance to salt stress (Hayashi et al., 1997). Another mechanism, Na + vacuolar compartmentalization, is dependent on the Na + /H + anti-porter, due to the ph change across the tonoplast membrane. AtNHX1 was isolated from Arabidopsis as a Na + /H + anti-porter similar to mammalian NHE transporters. When 6

24 this gene was over-expressed, more transporters were found in the tonoplast and salt tolerance was increased (Apse et al., 1999). Eight other AtNHX loci were also cloned in following studies and some of them were shown to be induced by hyper-osmotic stress and this response is dependent upon the hormone ABA. Recently, it was reported that Ca 2+ also plays an important role in osmotic signaling triggered by cold, drought and salinity, suggesting that the calcium sensor signaling network can induce specific stress responses to improve plant survival under saline conditions (Boudsocq 2010). Thus, the mechanisms against osmotic stress and osmotic stress induced cell signaling pathways can be considered in the study of salt stress Dehydration (drought stress) Drought is another important environmental stress affecting crop yields and qualities. As mentioned above, high chemical concentrations surrounding plants cause a water potential imbalance, resulting in dehydration ( micro-drought ). It was found that when water potential difference is greater than turgor loss caused by salt chemicals, cellular dehydration happens. Studies of both leaves (Passioura and Munns, 2000) and roots (Rodríguez et al., 1997) suggested that the rapid and transient reductions in expansion or growth rate followed by a rapid and sudden increase in salinity are due to changes in cell water deprivation; roots had a much better growth recovery compared with shoot (Hsiao and Xu, 2000). Cellular responses of plants during drought stress include roots becoming thicker to penetrate compacted soil layers (Pathan et al., 2004), stomata closing to reduce water loss (Trejo and Davies 1991) and also reduction of carbohydrate metabolism (Keller and Ludlow 1993). Because drought is also caused by turgor loss, similar to osmotic stress, the synthesis of osmolytes and osmoprotectants is also one mechanism for plants to tolerate a water deficit. All these responses and mechanisms, to some extent, are 7

25 also involved in salt stress. The most common properties shared by salt stress and drought are responsive cell signaling transduction. It was found that genes responsive to dehydration are also responsive to high salinity, such as RD29A and RD29B, which are now usually used as positive control genes for salt and drought responses (Bartels and Sunkar 2005). The ABA independent regulatory element Dehydration-responsive element/c-repeat (DRE/CRT) also functions in high-salt-responsive gene expression (Yamaguchi-Shinozaki and Shinozaki, 2005). ABA, which is an important hormone involved in plant growth and development, is very important in environmental stress regulation (especially drought and salt stress). This point will be further discussed in the next part Ion disequilibrium Ion homeostasis is necessary for a plant to provide the optimum conditions for enzyme activity, to maintain the turgor pressure around particular values and also to be an important component in signaling. However, high salinity stress can break ion homeostasis, causing ionic stress that is specific to salt stress. A high level of Na + is toxic to plants because it interferes with K + nutrition and thus affects K + stimulated enzyme activities, metabolism and photosynthesis. First, the excessive amount of NaCl will lead to a competition between Na + and K + transport into cells due to their similar chemical properties, which induces the loss of K + /Na + balance (Rubio et al., 1995). Second, it was reported that K + is important in maintaining the activities of many enzymes inside the cell, while the excessive Na + will cause toxicity to many enzymes. Therefore, the ratio of K + /Na + contributes to the ability of plants to tolerate salt stress (Shabala and Cuin, 2008; Luan et al., 2009). 8

26 Oxidative stress It is known that drought, salt and cold stress can induce the accumulation of ROS (Abogadallah 2010). These include superoxide, hydrogen peroxide, and hydroxyl radicals. While a good effect of ROS is that they can induce ROS scavengers and some protective mechanisms, like ABA mediated pathway and osmotic adjustment (Jithesh et al., 2006), excessive ROS can have damaging effects on cellular structures and macromolecules such as lipids, enzymes and DNA (reviewed in Abogadallah, 2010), resulting in oxidative stress. In plants, there exist several strategies against oxidative stress, such as reduction of photosynthesis and anti-oxidative responses. ROS is mainly produced through photosynthesis, photorespiration and respiration, as well as extra oxidases such as NADPH oxidases, and amine oxidases etc. (Guzy et al., 2005). Study in cyanobacteria suggested that salt stress can inhibit photosystems II and I for reducing the oxidative stress (Allakhverdiev and Murata, 2008). In addition, producing a number of antioxidants in plants is a very important strategy for ROS homeostasis, reducing the bad effect and it has been shown that high levels of antioxidants in plants can help resist oxidative damage (Spychalla and Desborough, 1990). These enzymatic pathway and antioxidant coding genes are discussed in detail in this review (Jithesh et al., 2006), including superoxide dismutase (SOD), catalases (CAT), ascorbate peroxidases (APX) and peroxidases. For example, loss of function of catalases in tobacco and Arabidopsis showed enhanced sensitivity to oxidative stress under salt conditions (Willekens et al., 1997; Cao et al., 2005). It was also shown that the putative phospholipid hydroperoxide glutathione peroxidase (PHGPX) transcript can be induced by oxidative stress and salt stress in Arabidopsis (Sugimoto and Sakamoto, 1997). Another mechanism to protect plants from oxidative stress is the accumulation of osmolytes. For example, proline and glycine betaine were reported to induce antioxidant defense gene expression and suppress cell death in cultured tobacco cells under salt stress (Banu et al., 2009). 9

27 1.1.4 Hormone involvement in salt response It is believed that hormones can mediate the conversion of developmental and environmental information into a cellular context by regulating a series of genes expression and biological processes. Although the exact roles of hormones involved in environmental stresses such as salt stress are not clear, the environmental stimuli often influence cellular concentrations of plant hormones (Ghanem et al., 2008) and the subsequent regulation of a series of genes, which lead plants to an ultimate adaptive condition. In addition, environmental stresses can affect different steps in a hormonal signaling pathway, including biosynthesis, perception (receptors), transport and downstream targets etc. Several major hormones will be discussed in the following sections with regards to their biosynthetic and signaling pathways, as well as their involvements in salt stress response or tolerance Abscisic acid (ABA) ABA is a phyto-hormone that is important in plant growth and development, as well as environmental stress which controls downstream stress responsive genes. First, ABA biosynthesis is affected or involved in salt stress response and tolerance. It has been shown that a high concentration of salinity increases ABA level, mainly due to the induction of gene expression for ABA biosynthetic enzymes (Xiong et al., 2002; Geng et al., 2013). Important genes including Zeathanxin epoxidase (ABA1), ABA2, 9-cisepoxycarotenoid (NCED), ABA aldehyde oxidase (AAO) and ABA3 have been cloned (Xiong et al., 2002). ABA functions in salt tolerance through regulating the downstream stress-responsive target genes and the corresponding physiological events such as stomata closure. Stomata are pores in the epidermis of leaves and stems used to control gas exchange and water 10

28 transpiration. CO 2 and O 2 enter the plant through stomata and are used by plants for photosynthesis and respiration (Farooq et al., 2009). Also, water evaporates through these pores. A stoma is surrounded by a pair of guard cells that control the size of the pore. When the turgor of the guard cells decreases, stomata close to prevent water from leaking out (Outlaw 2003). Mechanistic studies indicated that ABA can target guard cells to induce stomata closure through Ca 2+ flow under drought and oxidative stress, as well as salt stress, reducing water loss or photosynthesis (Chaves et al., 2009). ABA regulated target responsive gene expression is another important mechanism for salt tolerance. It is known that stress responsive genes expression is regulated either through an ABAdependent or ABA-independent pathway. ABRE (ABA-responsive element, PyACGTGGC) is the cis-acting element mediating ABA induced gene expression. Genes, such as RD29B and RD20A, have this element in their promoter regions and the removal of this element affects the induction of gene expression under ABA or stress (Yamaguchi- Shinozaki and Shinozaki, 2005). ABRE interacts with the bzip transcription factors AREBs/ABFs, which can be induced by ABA signaling. Also important is the induction of MYB2 and MYC2 transcription factors that regulate genes containing MYB and MYC binding motifs (C/TAACNA/G, and CANNTG), such as RD22. ABA-independent pathways involved in drought and salt stress are mainly mediated by DRE/CRT (drought responsive element, A/GCCGAC). For example, the drought and salt induced RD29A is a gene containing this element. It was shown that the AP2/ERF family transcription factors can be induced by stress signals (perception) and bind to these elements and activate gene expression. NAC, HD-ZIP transcription factors, are also involved in the ABAindependent pathway in stress responses (Yamaguchi-Shinozaki and Shinozaki, 2005). Since the organism is very complicated, cross-talk between these pathways must exist; for example, the regulation of the gene RD29A is both ABA-dependent and ABA- 11

29 independent, and proline accumulation for osmotic stress can be mediated by both the pathways (Savouréet al., 1997). Fluridone (1-methyl-3-phenyl-5-(3-trifluromethyl (phenyl))-4-(1h)-pyridinone) is an herbicide whose mode of action at the molecular level has not been clearly elucidated, but it has been widely used in the study of ABA functions involved in many biological processes as an ABA biosynthetic inhibitor, probably by inhibiting formation of carotenoids that are the main precursors for ABA synthesis in plants (Zeevaart and Creelman, 1988). Many studies have used this chemical to block ABA biosynthesis, although it is not clear to what extent the blocking occurs. For example, a recent study showed that fluridone can promote the division of stem cells in the quiescent center by inhibiting ABA s function, because exogenous ABA suppressed the QC cell division (Zhang et al., 2010). It was also used to study the role of ABA in lateral root development; exogenous ABA inhibits lateral root initiation and emergence at concentrations of 1μM or the above, but this inhibition can be released by fluridone at the same concentration (Hooker and Thorpe, 1998) Ethylene Ethylene (C 2 H 4 ) is a very important gaseous hormone, participating in stress response, as well as many other developmental processes, such as germination, fruit ripening, and organ abscission etc. Although ethylene can be produced in all tissues of plants, variants still exist in different tissue types, different developmental stages and specific environmental conditions. The substrate for ethylene biosynthesis is the amino acid methionine, and the important enzymes involved in biosynthesis are AdoMet synthetase, ACC synthases and ACC oxidase (ACO). ACC is a critical precursor of ethylene and it is often used as a method of ethylene treatment since it is difficult to control the amount and 12

30 concentration of ethylene gas. The formation of ACC is a rate-limiting step in ethylene biosynthesis, and the ACC synthase (ACS) genes have been cloned (ACS1-13, Sato and Theologis, 1989). The effects of ethylene in salt stress response lies in different levels. First, the effect of salt stress can be imposed on the first step of ethylene signaling perception. According to sequence similarity and structural characteristics, ethylene receptors can be divided into two groups, I and II. In Arabidopsis, ETR1 and ERS1 belong to group I, and ETR2, EIN4 and ERS2 belong to group II (Cao et al., 2008). Zhao et al. (2004) found that the expression of ETR1 is down-regulated by salt and osmotic stress at both transcription and translation levels in Arabidopsis. Transgenic tobacco plants over-expressing the group II ethylene receptor NTHK1 gene showed higher sensitivity to salt stress compared with wild type (Cao et al., 2006). The effects of ethylene receptors in salt stress to some extent are to regulate downstream salt-responsive gene expression, such as AtERF4, RD21A, AtNAC2, and BBC1 etc (reviewed in Cao et al., 2008). Other components of ethylene signaling can also be involved in salt stress response. CTR1 is a negative regulator to ethylene signaling downstream of ETR1. The ctr1-1 mutant showed increased salt tolerance and the germination rate and development of this mutant are better under salt and osmotic treatment (Achard et al., 2006). Another effect of ethylene in salt stress is its interaction with other hormones, such as ABA. It has been shown that ethylene level can be reduced by ABA under salt stress, resulting in reduction of leaf abscission probably by decreasing the accumulation of toxic Cl - ions in leaves. It was also shown that disruption of EIN2, which is a central factor of the ethylene signaling pathway in plants, changed the expression pattern of RD29B under salt stress, which is regulated in an ABA-dependent pathway (Wang et al., 2007). 13

31 Gibberellic acid (GA3) GA3 is another plant hormone which is a positive regulator of growth and development. It has been found in Arabidopsis that GA3 participates in many events, such as seed germination, leaf and root growth, inflorescence stem elongation, anther/petal development, and fruit/seed development and so on (reviewed Schwechheimer 2008). The biosynthesis of GA in higher plants has been studied clearly (reviewed in Sun 2008). GID1 is GA receptor first identified in rice, the loss of function of which can lead to a dwarf phenotype (Ueguchi-Tanaka et al., 2005); in Arabidopsis there are three orthologs GID1a, b and c. The triple mutant in Arabidopsis showed failure in flower development. DELLA proteins are a subfamily of plant-specific GRAS (GAI, RGA and SCARECROW) family, and they function negatively in plant growth. GAs may promote plant growth through binding and degrading DELLA proteins (Wen and Chang, 2002), so in some studies DELLA proteins were used as an indicator for the change of GAs. According to previous studies, it was known that GA participates in the stress response, including salt stress. On the one hand, the biosynthesis of GA can be affected by salt stress. For example, it was reported that high salinity greatly represses GA3 oxidase1 (GA3ox1) gene expression (Kim et al., 2008). Growth decrease induced by salt stress may be via modulating the GA metabolic pathway; because it was found that salt-treated Arabidopsis plants contain reduced levels of bioactive GAs (Achard et al., 2006). The most famous evidence is the study of salinity-responsive DDF1, which encodes an AP2 transcription factor of the DREB1/CBF (drought responsive element binding protein) subfamily. Overexpression of this transcription factor can reduce GA levels and at the same time increase salt tolerance (Achard et al., 2006). Also, in rice, GA3 reduces NaClinhibition of seed germination through enhancing hydrolysis of starch in endosperm (Lin and Kao 1995). Independent of ABA, the GA pathway mediates the salt regulation of seed germination through a membrane-bound NAC transcription factor NTL8, which is 14

32 induced by high salinity. The germination of ntl8 mutant seeds is resistant to high salinity and PAC (GA biosynthesis inhibitor), suggesting NTL8 modulates GA-mediated salt signaling in regulating seed germination (Kim et al., 2008). PAC (Paclobutrazol) is an inhibitor of GA synthesis that inhibits mono-oxygenases involved in converting ent-kaurene to ent-kaurenoic acid. It has been widely used for countless studies for the functions of GAs in regulating biological processes Brassinosteroids Brassinosteroids (BRs), also referred as brassinolide (BL) are a group of steroidal plant hormones that play essential roles in a wide range of developmental phenomena and environmental stress responses (Khripach et al., 1998). BRs are synthesized from phytosterol precursors that differ from each other by their aliphatic substituents at the C- 24 position, and the biosynthetic pathway is reviewed by Fujioka and Yokota (2003). Forward genetics has isolated BR-deficient mutants, in which the mutated genes characterized are involved in BR biosynthesis. DET2, SAX1, DWF4, and CPD genes are involved in different steps during the biosynthesis, and their mutations all cause a strong dwarf phenotype. Studies have focused on the identification of BR signaling (details are reviewed in Wang et al., 2012). BRs are reportedly involved in different environmental stresses including salt stress. BR functions in high salinity stress through two potential mechanisms. One is to protect plants from oxidative damage (Schutzendubel and Poll 2002). The exogenous application of BRs can effectively reduce the adverse effects of salt stress or induce salt tolerance, such as overcoming inhibition of seed germination by salt (Kagale et al., 2007, Ali et al., 2008). This is potentially by modifying the activities of important antioxidant enzymes (Shahbaz et al., 2008). Also, when treated with HBL, the salt induced high level of H 2 O 2 15

33 and lipid peroxidation is reduced, and this effect is at a transcriptional level (Cao et al., 2005). The other mechanism that BR uses is to eliminate ion disequilibrium. For example, BRs have been found to improve the Ca 2+ /Na + and K + /Na + ratios of wheat cultivars by enhancing Ca 2+ and K + uptake, and reducing Na+ uptake, which may have contributed to enhanced salt tolerance (Qasim et al. 2006). The cross-talk between BRs and ABA or other hormones can also be a mechanism for eliminating ion disequilibrium Cytokinin Cytokinin, named after its function in promoting cell division, has been found in all higher plants. The most common form of naturally occurring cytokinin in plants today is zeatin that was isolated from Zea mays (Letham 1963). Cytokinin plays a vital role in regulating cell proliferation and organ differentiation, and it is especially active in the meristematic region. Also, it is an important regulator of growth and enlargement of root/shoot and leaves (reviewed in Sakakibara 2006). The study on cytokinin-deficient plants suggested that the regulatory functions of cytokinin in root and shoot meristems are opposite (Werner et al., 2003). It showed that cytokinin is required in the growth of shoot apical meristems and leaf primordial, while it is a negative regulator of root growth and lateral root formation through controlling the exit of cells from the root meristem. Cytokinin biosynthesis occurs through the biochemical modification of adenine. ip (isopentenyladenine) and trans-zeatin mainly originate from the methylerythritol phosphate pathway (MEP ) and most of the cis-zeatin is derived from the MVA (mevalonic acid) pathway (reviewed in Sakakibara 2006). IPT (adenosine phosphateisopentenyltransferase) genes have been characterized in Arabidopsis and studied widely, including adenosine phosphate-isopentenyltransferase, which catalyzes the first step of 16

34 biosynthesis and genes coding for the subsequent steps, such as CYP735A1 and CYP735A2 (reviewed in Sakakibara 2006). It was reported that the levels of ipa (isopentenyladenosine) and ZR (zeatin riboside) were greatly induced by salinity stress in maize and pea (Atanassova et al.,1997). Similarly, change in the cytokinin content in plants induced by salt stress was frequently reported. For example, the decrease of active isoprenoid cytokinin level was observed in barley roots and shoots, which happens after treatment with a high concentration of NaCl (Kuiper et al., 1990). This indicated that cytokinin plays important roles in the salt stress response to maintain growth of plants under stress or to maintain cross-talk with other hormones for salt tolerance. However, the exact role or mechanism how cytokinin functions in salt stress is not known, and few studies have tried to explore it. For instance, a microarray analysis of Arabidopsis CK receptor mutants showed that CK signaling can be involved in salt stress response by up-regulating many stress-inducible genes (Tran et al., 2007). Another study on maize showed that salt and osmotic stresses induces expression of some CK biosynthetic genes while genes involved in CK signal transduction are uniformly down-regulated (Vyroubalova et al., 2009) Auxin Auxin was first studied in late 19th century, described as an "influence" that could move from the tip of the coleoptile to the lower region where it controlled bending. IAA (indole-3-acetic acid) is the most important auxin produced by plants, although other natural forms exist, such as IBA (indole-3-butyric acid). Since auxin biosynthesis in vivo is extremely complex, there is not a confirmed pathway (tryptophan-dependent and - independent pathways), but recent studies contributed in finding out some important genes for the tryptophan-dependent pathways. Cheng et al. (2007) found that YUCCA 17

35 family proteins are important for auxin production, because the over-expression of YUCCA genes showed an auxin-overproduction phenotype and the yuc quadruple mutant failed in establishing a basal-apical axis in embryogenesis and normal development of root meristem. Also, based on mutant analysis, the ethylene responsive gene TAA1 was found to encode an aminotransferase catalyzing the conversion of tryptophan to IPA (Stepanova et al., 2008), suggesting another biosynthesis pathway cross talking to ethylene. With the importance of auxin, synthetic auxins were made as supplements, such as NAA (naphthaleneacetic acid) and 2, 4-D (2, 4-dichlorophenoxy-acetic acid) etc. Auxin plays important roles in almost all aspects, including development of embryo (apical-basal formation), leaf and root formation and development, phototropism and gravitropism, ethylene biosynthesis, as well as in the regeneration etc. Among these functions, responses to environmental stimuli are important, such as phototropism, gravitropism and wounding induced regeneration. However, for auxin s involvement in salt stress, there are limited studies, though some clues indicate that auxin can function in salt stress, mainly as a co-factor in the other signaling pathways such as ethylene and salicylic acid (reviewed in Gavan-Ampudia and Testerink, 2011) Studies of high salinity stress in Arabidopsis Arabidopsis is a model plant in salt stress studies As mentioned above, plants can be categorized as halophytes and glycophytes based on their capacity to tolerate salt in the environment. Halophytes can grow on a salt concentration as high as over 400 mm, which is about 10 times that of glycophytes (Flowers et al., 1977). Halophytes have much higher water use efficiency, low internal 18

36 carbon dioxide concentration, efficient solute accumulation, and low levels of Na + and Cl - ions in the cytoplasm and chloroplast. However, only 2% of the terrestrial plants are halophytes and the majority of crops are glycophytes, so more effort has been made by scientists and breeders to study glycophytes in order to engineer salt tolerant crops (Bohnert et al., 1995). It was believed that glycophytes and halophytes have the same or similar salt tolerance machinery, which may not be operating effectively in normal conditions for glycophytes. Due to the complexity of salt tolerance mechanism, a good genetic model is necessary. Arabidopsis thaliana has been used as a model plant for many studies, including salt response and tolerance. As a genetic model, Arabidopsis thaliana has desirable life history traits, such as short life cycle, self-pollination and high seed number. Also, it has a small genome and its genomic background is easily accessed, such as transcriptomes under different conditions and developmental stages, proteome, and epigenome, so correlation of regulation at different levels can be analyzed. In addition, genetic manipulation is easy, such as efficient and stable transgenic integration (inflorescence dipping method), mutagenesis, and mutant screening, which are indispensable for functional studies in mechanism discovery. Many gene knock-out lines and RNA or DNA arrays are easy to obtain commercially. The most important aspect is that Arabidopsis is a glycophyte sensitive to high salt concentration, so the salt tolerance mechanism revealed in Arabidopsis can be used in other corps. Due to all of the above advantages, Arabidopsis has been an ideal model for salt stress studies. A number of genes involved in salt response and tolerance mechanisms have been identified, characterized and cloned from Arabidopsis, among which the most important are the Salt Overly Sensitive (SOS) loci (Zhu, 2000). Approximately 250,000 mutagenized seedlings were screened using a root-bending assay (Wu et al., 1996; Zhu et 19

37 al., 1998), and 5 genes were characterized as salt tolerance genes because the mutants of these genes showed salt hypersensitive properties. SOS1, encodes a putative Na + /H + antitransporter with a molecular mass of 127 kd and its transcript is up-regulated by NaCl stress (Shi et al., 2000). SOS2 gene encodes a Ser/Thr protein kinase with an estimated molecular mass of 51 kd (Liu et al., 2000). Mutational study suggested that the C- terminal regulatory domain of this gene is essential for its protein function in plant salt tolerance (Liu et al., 2000). SOS3, encodes a Ca 2+ binding protein with three predicted EF-hands (Liu and Zhu, 1998). SOS4, encodes a pyridoxal kinase that is involved in the biosynthesis of pyridoxal-5-phosphate, an active form of vitamin B6, which might regulate Na + and K + homeostasis by modulating the activities of ion transporters (Shi et al., 2002). SOS5 encodes a putative cell surface adhesion protein and is required for normal cell expansion. Under salt stress, the root tips of sos5 mutant plants swell and root growth is arrested and this phenotype is caused by abnormal expansion of epidermal, cortical and endodermal cells controlled by cell-to-cell adhesion in plants (Shi et al., 2003). SOS1, SOS2 and SOS3 are in the same salt tolerance pathway for Na + homeostasis (Zhu, 2000). To illustrate this pathway, high Na + concentration stress leads to an increase of cytosolic free Ca 2+, which binds with SOS3. Then activated SOS3 can activate SOS2 kinase, and this complex further positively regulates SOS1, which exports Na + from the cell and maintains the homeostasis of Na + and K + (reviewed in Xiong and Zhu, 2002) Root a multicellular organ directly responsive to salt stress Plants survive using the water and nutrients absorbed and transported by root from soil or growth medium. So the root is the organ directly interacting with high salinity and the mechanisms developed in root for salt perception, response, tolerance and adaption were focused on by generations of scientists (Drew and Lynch 1980). It has been shown that 20

38 sodium ions enter the root cell passively through two kinds of cation channels; voltagedependent cation channels and voltage-independent channels (VIC) (White, 1999; Amtmann and Sanders, 1999). K + transporter HKT1 was first identified in a study of wheat roots. It was found that when the external Na + concentration increased, the HKT1 could function as a low affinity Na + channel, leading to Na + influx and the hkt1 mutant showed a lower Na + content in plants (Rubio et al., 1995). The mechanism of VIC is not very clear, though a cyclic nucleotide-based signaling pathway may affect Na + transport via VICs. On the other hand, roots also develop different strategies against accumulation of Na + in the cell by exclusion of Na + ions by SOS pathway mediated by Ca 2+ ions, compartmentalization of ions at the cellular level, and induction of anti-oxidative enzymes and hormones etc. However, the root is composed of a series of cells with different identities (Benfey and Scheres 2000); the roles played by these different cells and signal transduction among these cells involved in salt response and adaption are not clear. Arabidopsis root is a good multicellular model for the study of development and environmental responses because of its simple but highly organized radial pattern (Figure 1). Along the radial axis, root is composed of several different cell types, such as epidermis in the outer layer, cortex, endodermis and stele (including xylem, phloem, and pericycle cell types). At the root tip there is a structure called root cap that is composed of two parts, lateral root cap (out of epidermis but terminating at the transition zone) and columella at the root tip. These cells are developed from their initial cells around the QC cells that are almost mitotically inactive (Figure 1). Along the longitudinal axis, it can be divided into 3 zones, meristem zone, elongation zone (there is a small transition region between meristem and elongation zones) and maturation zone. Cells in meristem zone are mitotically dividing from the initial cells. When cells enter elongation zone, they became 21

39 elongated and differentiated. When in the maturation zone, the lateral organ, lateral root developed from the root cells. Previous studies suggested that salt stress can cause specific effects to different cell types. For example, the root hairs specifically developing from epidermal cells are inhibited mildly and the identities of H/N epidermis can be affected under salt stress (Halperin et al., 2003); cortical cell swelling could be also caused by salt stress (Burssens et al., 2000); lateral root development from pericycle is also inhibited by high concentrations of salt (Duan et al., 2013). However, there are few studies systematically focused on why salt stress can cause these cell type specific changes or that study salt stress with respect to cell type specificity in Arabidopsis and rice (Plett et al., 2010; Kiegle et al., 2000; Ma et al., 2007). The first spatial transcriptional map of Arabidopsis root under salt stress for 1 hour was generated by Dinneny et al. (2008), suggesting the transcriptional responses to environmental stresses leading to biological functions are mediated by developmental parameters and cell identities. For example, the specific repression of cell shape genes, such as COBRA and RSW3, in cortex and epidermis are well correlated to the radial swelling of the outer tissue layers caused by salt stress. 22

40 Figure 1. Schematic longitudinal and cross section of Arabidopsis root tip (Adapted and modified from Dinneny et al., 2008). The structure of Arabidopsis root is composed stem cell niche and radially organized cell files. The quiescent cells (dark blue) are stem cells that divide into stem cells and cells that can differentiate into cells with different identities. The radially organized cell files include epidermis and lateral root cap (pink), cortex (yellow), endodermis (green), and stele (purple) in which there are phloem (red) and xylem. In the root tip is columella. 23

41 1.2 Transcriptional regulation and transcriptional network Transcription, which is the first step of genetic information flow, is indispensable in converting developmental and environmental cues into biological consequences. In most organisms, many of the biological processes are regulated at the level of transcription, such as the determination of polarity during embryogenesis in Drosophila (Lilly et al., 1994), organogenesis and differentiation in plants (Weigel and Meyerowltz 1994; Helariutta et al., 2000). Transcriptional regulation of gene expression is also widely involved in response to environmental cues and stresses, as well as the defense response against pathogens (reviewed in Scott 2000; Benfey and Weigel 2001). In eukaryotes, the large number and variety of transcription factors as well as the combinatorial property of CREs determine the diversification of gene expression patterns, which is required for biological complexity (Struhl 1999). The availability of genome sequences and the development of systematic approaches make it possible to compile the complex gene expression regulation into a network, showing the flow of information throughout a biological system and prediction of how transcript or protein expression will change in response to a certain stimulus or at a specific developmental time (Barabasi and Oltvai 2004; Alon 2007) Transcriptional regulation is an indispensable process involved in developmental process and environmental stimuli response A large portion of the functional proteins in a genome are transcription factors (TF). For example, transcription regulators represent approximately 4.6, 3.5, 3.5, and 6% of the genes in Drosophila, C. elegans, yeast and Arabidopsis (Riechmann et al., 2000). The TFs encoded by the A. thaliana genome can be classified into more than 40 major 24

42 families (Riechmann 2006). The three largest are the MYB superfamily, the basic helixloop-helix (bhlh) family, and the AP2/EREBP family, all having more than 120 members (Riechmann 2006; Shiu et al., 2005). TFs that cannot be assigned into any TF family appear to be rare, but they also have very important developmental functions, such as the floral meristem identity gene LFY and SPL/NZL. The great number and diversity of TFs determines their ability to regulate complex patterns, and more importantly, their wide function for development and response to environmental stress. Transcriptional regulation is decisive for many developmental events, such as polarity or cell fate decision during embryogenesis, and the formation of organs. The anteriorposterior axis polarity decision during Drosophila embryogenesis involves several transcription factors: Bicoid, Hunchback and Caudal (reviewed in Dearden and Akam 1999). An embryo from a bicoid mutant mother develops a lethal phenotype with posterior structures on both ends (reviewed in Johnston and Nusslein-Volhard 1992). These TFs were also reported to transcriptionally regulate other developmental genes, such as gap genes, whose patterns are located in different segments of the embryo along the anterior-posterior axis (Ephrussi and Johnston 2004). In plants, flower development is a classic example involving transcriptional regulation in the whole process, from the formation of the floral meristem to the normal development of flower structure. WUS, a plant-specific homeobox family transcription factor, is a master regulator for the maintenance of shoot meristem and formation of floral meristem (Mayer et al., 1996). Other WOX genes in the same family are also involved in this process, and WOX5, at the same time, is a root apical meristem regulator (Graaff et al., 2009). The interactions between MDS-domain transcription factors result in the correct structure of the flower, which is summarized in the ABC model (Soltis et al., 2007). In addition, the productive transition is controlled through regulation of a group of genes of which LEAFY (LFY) is the most important one, encoding a plant-specific transcription factor 25

43 that can act as either an activator or repressor depending on what co-factors it is interacting with (Siriwardana and Lamb 2012). Other processes, such as the development of root and stomata, are also regulated transcriptionally by SHR/SCR and trihelix family TFs, such as GTL1 (Pascuzzi and Benfey 2009; Kaplan-Levy et al., 2012). Many transcription factors and their binding motifs have been reported involved in environmental stresses, such as the ERF/AP2 family transcription factors DREB2A and DREB1A/CBF3 (DRE/CRT as the binding motif), bzip family factors AREBs/ABFs (ABRE as the binding motif), MYCs (bhlh) / MYBs (MYB) (MYC/MYB binding motifs) and WRKYs (W-box) etc. It was recently reported that the transcription factors WRKY18, WRKY40 and WRKY60 can form a small network in the regulation of stress through ABA signaling because of their induction by ABA (Chen et al., 2010). Several core transcription factors, TOC1, LHY, and CCA1 are also involved in circadian clock (Gendron et al., 2012). The availability of the Arabidopsis genome sequence and computational methods allow a global, systematic genomic analysis of transcriptional regulation in plants, including environmental study. One of the interesting studies is the circadian clock response (Michael et al., 2008). In this study, the authors developed their pipelines based on the genome-scale data and module analysis, determining three different modules, GBOX, GATA, and PBX/TBX/SBX, respectively correlated to genes having peak expression in early morning, morning and afternoon. In addition, studies have shown that in drought and salt stress, ABA dependent and independent pathways are involved (elements of ABRE or DRE, Yamaguch-Shinozaki and Shinozaki, 2005) based on just a small portion of salt/drought responsive genes, such as RD29A Mechanisms of transcriptional regulation 26

44 Chromosomes, in which genes are localized, are normally in stable condensed packed conditions. So the process of transcription needs a series of proteins to switch from the stable chromatin to the active open DNA strands and transcribe into mrna, especially for induced gene expression. Briefly, the proteins involved in transcription in eukaryotes can be classified into the following functional groups (reviewed in Riechmann 2002). The first category is the basic transcription apparatus and intrinsic associated factors (general transcription factors; GTFs), including RNA polymerases, and the GTFs TFIIA, TFIIB, TFIID, TFIIE, TFIIF and TFIIH. The GTFs are necessary to aid the RNA polymerase to recognize and anchor on the promoter and unwind the DNA helix. The second category is sequence-specific DNA binding TFs, activators or repressors. They are for the selective gene expression regulation, and their expression is usually restricted to a specific tissue/cell type, a specific time, or a specific stimulus-dependent condition. These TFs usually bind to a specific cis-regulatory element (CRE) in the promoter that is distant from the core promoter (TATA box), through their DNA binding domains. The regulatory domain can cause conformational change of the DNA and facilitate the access of the basic transcription apparatus to the DNA (activator). Alternatively, the TFs can compete with an activator through binding the same regulatory element sequence to inhibit expression (repressor). The third category is large multi-subunit co-activators and other cofactors. This class of proteins can interact with sequence-specific TFs, modulating their binding or interaction with the core machinery. For example, the TOC1 is a co-factor of CHE (Pruneda-Paz et al., 2009) that regulates the expression of CCA1 in the evening in the circadian phenomenon. The last category of proteins is chromatin-related proteins. This group includes factors that covalently modify histones and remodeling complexes that hydrolyze ATP for reorganizing chromatin structure. Histone acetylation is generally a characteristic of transcribed chromatin, whereas deacetylation is associated with 27

45 repression. In addition, DNA methylation is also related to gene expression regulation because methylated DNA in the promoter can inhibit transcription. The process of transcriptional activation can be summarized: the activator is phosphorylated and enters the nucleus, then the enhancer upstream of the promoter is recognized by the activator. This activator recruits the SWI/SNF complex, resulting in the remodeling of chromatin and localized histone acetylation, which facilitates the access of additional transcription factors to cis-regulatory elements. These secondary activators aid gene transcription also through co-factors, recruiting the RNA polymerase complex (GTFs as well) to the transcription initiation site. The order of the proteins recruited can be different among promoters and organisms, but the roles of activators in regulation can be the same among eukaryotes including plants, that are enlisting chromatin modifying activities and then inducing localization of the basal transcription apparatus (Cosma et al., 2001; Brown et al., 2001; Merika and Thanos 2001) Approaches to generate a transcriptional network Since biological systems are very complex, the study of regulation for biological processes is difficult to elucidate. However, it can be simplified into multiple subsystems at the molecular level, so that one can determine how the cell and even the organism will behave based on the dynamic interactions and/or associations between the biomolecules. A network is a good way of integrating the complex associations and/or interactions where the nodes are biomolecules (DNA/genes, RNA/transcripts, proteins and metabolites) and the edges represent interactions and/or associations between them. In a transcriptional network, the nodes are regulatory proteins, such as transcription factors and co-factors, and their target regulated genes (proteins) (Barabasi and Oltvai 2004). 28

46 The first and simplest function of a transcriptional network is to tell how the regulated information flows for a biological process. It was summarized in studies on the global transcriptional regulatory systems in E. coli, revealing how environmental inputs were interpreted into the final output responses of gene expression. It can be found from the network that half of the endogenous TFs control the expression of exogenous and hybrid regulators while little reciprocal regulation from external TFs to internal TFs occurs, suggesting the external signals must be combined with internal signals in the network. In Arabidopsis, more and more transcriptional networks are generated. For example, a global gene regulatory network was generated to understand the differentiation of root epidermis cells, that is, cell fate for root hair and non-root hair cell (Bruex et al., 2012). Another simple transcriptional network was summarized on how the GRAS family transcription factors regulated stem cell niche specification and ground tissue patterning (Pascuzzi and Benfey 2009). In addition, networks were also used to understand the regulation of the plant iron deficiency response (Hindt and Guerinot 2012). Another function that a transcriptional network can reveal is how transcriptional regulation is processed between transcription factors, negative auto-regulation, positive auto-regulation, feed forward loops and feedback loops (Alon 2007). The stem cell maintenance regulator, WUS, which is expressed in L3 layer of the central zone of SAM (shoot apical meristem), promotes the expression of CLV3 in the L2 and L1 layers. In-turn, CLV3 inhibits the expression of WUS in the two layers. This feedback loop is very important in maintaining a constant cell number in the stem cell niche (Schoof et al., 2000). One way of generating a transcriptional network is to integrate the previously validated simple regulations together in a gradual process. This kind of network is very simple and clear, involving several core transcription factors and target genes based on experimental data. For example, in circadian clock, CCA1/LHY and TOC1 are expressed in the morning and evening, respectively. Then an association was found that CCA1/LHY can bind to 29

47 TOC1 promoter and inhibit its expression, and then a further association was found that TOC in-turn promotes CCA1 expression with the assistance of CHE (Pruneda-Paz et al., 2009). The other way of generating a transcriptional network is through analysis of genomic data, such as microarray and ChIP on chip. Brady et al. (2011) generated a gene regulatory network that determines stele-specific TFs and mirna expression based on the high resolution gene expression data and Y1H (yeast one hybrid) and Y2H (yeast two hybrid) assays. 1.3 Objectives and significance of this study There are two important objectives for this study. As described above, hormones play important roles in salt stress; however, how hormone signaling systematically integrates into the salt stress response is unknown. So the first objective is to reveal the functions of hormones in mediating spatiotemporal transcriptional regulation in salt stress response, and their functions involved in biological processes such as root growth. The other important objective for this study is to uncover the biological functions of cis-regulatory elements involved in transcriptional regulation in salt stress response, further setting up a CRE centered transcriptional network. The significance of this study lies in mainly three aspects. Firstly, the analysis of the spatiotemporal transcriptional map provides a basis for understanding how biological processes are regulated in environmental stresses (salt stress as a model), such as the timing when casparian strip has the strongest function. This can help the practical application of promoting the properties of plants defending the unpleasant environmental condition such as salt stress. Second, using bioinformatics analysis, this is the first time salt stress and different hormones responses are integrated, providing how hormones are spatiotemporally involved in salt stress response and the targets mediating these 30

48 connections. Third, I demonstrate a synthetic strategy to efficiently and authentically validate the function of cis-regulatory elements, which is essential to understand the mechanism of transcriptional regulation in salt stress response. 31

49 Chapter 2 Materials and methods 32

50 2.1 Plant materials Arabidopsis thaliana ecotype Columbia (Col-0) was used in generating all transgenic plants. The mutant aba2-sail (SAIL_407_E12) and the following transgenic lines, UAS:: abi1-1 (Duan et al., 2013), RGA::GFP:RGA (Achard et al., 2006), and CASP1::CASP1:GFP (Roppolo et al., 2011) are in Col-0 background. GAL4-VP16/UAS enhancer trap lines J0951, J0571, Q2500, and Q0990 (Haseloff 1998) are in C24 background. 2.2 Plant growth conditions and stress treatment Dry seeds were collected and dried out for at least one week before seed germination. Seeds were surface sterilized with the following procedures: washing with 95% EtOH solution for 5 minutes, shaking in 20% commercial Clorox bleach plus 0.1% Tween-20 solution for 5 minutes, and rinsing in sterile dh 2 O four times. The sterilized seeds were cold treated at 4 C. Then the seeds were germinated and vertically grown for 5 days on sterile standard condition, which is 1% Agar medium containing 1x Murashige and Skoog salt nutrients (MSP01-50LT, Caisson), 1% sucrose and 0.5 g/l MES hydrate, ph 5.7 (adjusted with KOH). For salt treatment and hormone treatment, the 5-day seedlings were transferred to standard condition supplemented with 140mM sodium chloride (NaCl, Sigma-Aldrich), Abscisic Acid (ABA, Sigma-Aldrich), Gibberellic Acid (GA, Sigma- Aldrich), Paclobutrazol (Sigma-Aldrich), and combination of sodium chloride and hormones. Germination and growth of seedlings were performed in a Percival CU41L4 incubator at a constant temperature of 22 C with long-day lighting conditions (16 hours light and 8 hours dark). To avoid the medium dry out, the plates were sealed with 2 layers of parafilm (Alcan Packaging). For q-rt PCR experiment, nylon mesh was used on the surface of medium for the purpose of transferring. 33

51 2.3 Generation of transgenic lines Sequences design of synthetic promoters The strategy of the synthetic promoter and design is introduced in in detail. The sequences of the multimerized units are listed in Table 1, for the synthetic promoters I have studied. The 35S minimal promoter used was the region of -90~+8 of cauliflower mosaic virus gene (Benfey and Chua 1990). And the RD29A minimal promoter used was the region of -54~+96 of RD29A gene (Narusaka et al., 2003). 34

52 Table 1. The multimerized unit sequences for synthetic promoters. Name Sequence* Data base Experiment ABRE_RD29A cagacgcttcatacgtgtccctttatctct spatiotemporal cluster 6 GUS/LUC; Y1H ABRE_ABI1 ttttcttcgtctacgtgtcgaccatccacc spatiotemporal cluster 6 GUS/LUC; Y1H ABRE_GmC_RD29A cagacgcttcatacgtctccctttatctct GUS/LUC; Y1H ABRE_GmT_ABI1 ttttcttcgtctacgtttcgaccatccacc GUS/LUC; Y1H ABRE_25bp_RD29A acgcttcatacgtgtccctttatct Y1H ABRE_25bp_ABI1 tcttcgtctacgtgtcgaccatcca Y1H DRE_RD29A aataaatatcataccgacatcagtttgaaa spatiotemporal GUS/LUC cluster 6 DRE_TIR1 aagccgcgataagccgaccccccctctcca spatiotemporal GUS/LUC cluster 6 MYC_At2g45180 tcttaagtcgctcacatgctattttatctc cluster 15 in Brady's data (cortex) GUS/LUC; Y1H MYC_At5g64620 acaaaaaagattcacatggtcactttactg cluster 15 in Brady's data (cortex) GUS/LUC; Y1H TELO_At5g61030 cggagaccacaaaaccctaaaagcaacaac cluster 20/6 in GUS/LUC; Brady's data (xylem); Y1H spatiotemporal cluster 9 TELO_At4g15770 ctaaatccctaaaaccctaaaaaaacacaa cluster 20/6 in Brady's data (xylem); spatiotemporal cluster 9 W-box_AT1G22500 gtacgatctgaaaagtcaaccatctttgta cluster 11 in Brady's data (epidermis); spatiotemporal cluster 10 W-box_At1g01750 gtgaggaaagaaaaagtcaaaaaaattgaa cluster 11 in Brady's data (epidermis); spatiotemporal cluster 10 L1box_AT2G39510 ttttttttttttaaatgtaaccacggtaac spatiotemporal cluster 17 L1box_AT3G26610 agttttgtatgtaaatgtagagaaacagaa spatiotemporal cluster 17 AACCACT_AT1G13600 cctcttacttcaaaaccacttcacacaaca spatiotemporal cluster 2 AACCACT_AT5G23920 cagagaaggttaaccactccatatctacac spatiotemporal cluster 2 ATATAAT_AT1G14280 aaccaagacaaaatataattaatattttgt spatiotemporal cluster 19 ATATAAT_AT5G53250 aatctttaattaatataatgacaacgcacc spatiotemporal cluster 19 *The bold letters represent the core sequences of the cis-regulatory elements. GUS/LUC; Y1H GUS/LUC; Y1H GUS/LUC; Y1H GUS GUS GUS/LUC; Y1H GUS/LUC; Y1H GUS/LUC; Y1H GUS/LUC; Y1H 35

53 2.3.2 Constructs For the synthetic promoter reporters: the cdna of GUS was cloned from pgreen-hy107 vector provided by Dr. Hao Yu s lab (primers are in Table 2), and the fragment was cloned into pentr/d-topo vector (Invitrogen). LUC_sp reporter in pentr/d-topo vector is from Dr. Philip Benfey s lab. The synthetic promoters, synthesized from Gene Art or Gene Script, are composed of the 3 or 6 repeats accordingly of the units (listed in Table 1), 35S or RD29A MP after the repetitive units, and the gateway sites (attb4/attr1) at the ends. Multisite Gateway recombination reaction was done using LR clonase kit (Invitrogen) for the generation of the constructs of synthetic promoter::gus/luc into the dpgreen-mcherry (mcherry as the selection marker for positive transgenic plants) destination binary vector. For TFs over-expression line: the cdna clones of GBF2, BZIP3, and AZF3 are in pentr/d-topo vector provided by our collaborator, Dr. Jose Pruneda-Paz s lab. The cdna of ABF3 and STZ are cloned from genomic cdna using the primers in Table 2, and also cloned into pentr/d-topo vector. Single site gate way recombination was done using the LR clonase kit (Invitrogen) to integrate the coding sequences of the TFs into the binary destination vector pubn-yfp-dest (Figure, bought from Dr. Christopher Grefen). 36

54 Table 2. Primer sequences used in cloning, sequencing and colony PCR. Primer Description name N-65 Forward_RD29A promoter for gateway N-66 Reverse_RD29A promoter for gateway N-122 RD29A promoter sequencing 1 N-123 RD29A promoter sequencing 2 N-1973 Forward_ABI1 promoter for gateway N-1974 Reverse_ABI1 promoter for gateway N-1983 ABI1 promotr sequencing 1 N-1984 ABI1 promotr sequencing 2 N-1294 Forward_cDNA of GUS N-1295 Reverse_cDNA of GUS N-1296 GUS cdna sequencing 1 N-1297 GUS cdna sequencing 2 N-2689 Forward_Luciferas e-colony PCR N-2690 Reverse_Luciferas e-colony PCR N-3043 Forward_cDNA of at4g34000 N-3044 Reverse_cDNA of at4g34000 N-3045 Forward_cDNA of at1g27730 N-3046 Reverse_cDNA of at1g27730 Sequence GGGGACAACTTTGTATAGAAAAGT TGGGAGATCTCAAAGTTTGAAAG GGGGACTGCTTTTTTGTACAAACTT GTTTCCAAAG ATTTTTTTCTTTCCA GGAGATCTCAAAGTTTGAAAG AGGAGAAATACAATTCGAATG GGGGACAACTTTGTATAGA AAAGTTGTTCGATGATTTC TCGCTCTTT GGGGACTGCTTTTTTGTACAAACTTGT AACGGTAAAGATTTGATCTTTTTC TCTTGCCCATCATCCAAAG ACCAACTCTTCATTTCCCT CACCCAGTCCCTTATGTTACGTCCT TCATTGTTTGCCTCCCTGCTGCG TTAACTATGCCGGAATCCAT TACCCTTACGCTGAAGAGA TCTATCCGCTGGAAGATGGA CCGTGCTCCAAAACAACAAC CACCATGGGGTCTAGATTAAACTT CAAGAG CTACCAGGGACCCGTCAATG CACCATGGCGCTCGAGGCTCTTAC TTAAAGTTGAAGTTTGACCGGAAAGT 37

55 2.3.3 Agrobacterium mediated plant transformation The frozen GV3101 competent cells were thawed on ice. 1-2 μl binary vectors with transgenes were added to 20 μl GV3101 cells on ice and mixed sufficiently by gentle tapping the tube. Then the mixtures were put into the 1 mm Gene Pulser cuvette (Bio-rad) and subject to electroporation at 1.8 kv, and then the electroporated bacteria were cultured in 500 μl LB liquid medium for 1 hour with shaking at 28 C and spread onto selective LB agar medium, containing 10 μg/ml Tetracycline, 100 μg/ml Spectinomycin and 50 μg/ml Gentamycine. Colony PCR was performed on the single colonies to verify the transgenes after incubating the plates under 28 C for 2 days. The single colonies were then subject to Agrobacterium-mediated floral dip method established previously (Clough and Bent, 1998) and some modifications for simplification were made in our lab. Specifically, the positive single colonies were then cultured on large petri dishes (150 mm diameter) of selective LB agar medium for 2 days at 28 C. Bacteria cells were scraped and collected into 200 ml infiltration medium (1/2 MS salt, 5% sucrose, with freshly added 0.03% Silwet L77, ph 5.7 adjusted with KOH). Arabidopsis plants were grown at 5-8 seedlings per pot. After bolting, the primary shoots were cut off and more secondary shoots were emerged after 7 days. Subsequently, the shoot part of the plants was dipped into the above prepared cell suspension medium (make sure all the inflorescences are dipped into the solution) for 30 seconds. Plant pots were put horizontally under dark for 24 hours, and then the plants were grown under normal growth condition in greenhouse. Seeds of T1 generation were collected and the positives were selected based on BASTA resistance or visually based on mcherry fluorescence using an M165 FC fluorescence microscope (Leica). 2.4 Yeast one hybrid screen 38

56 2.4.1 Constructs generation The synthetic promoters, synthesized by Gene Art (Life Technologies), are composed of the 3 or 6 repeats of the units (listed in Table 1) and include the overhang for golden gate cloning at the ends. The flanking sequences for golden gate cloning are: GGTCTCAAGTA at the 5 and ATCTAGAGACC at the 3. The dpgreen-mcherry binary vector for golden gate cloning was modified from the version of gateway dpgreenmcherry vector, and the overhang after BsaI cutting is tcat-3 and 5 -tgga Yeast transformation The protocol of yeast transformation in our lab is a modified version of the previously published high-efficiency yeast transformation method (Deplancke et al., 2004). In brief, the constructs were first linearized using the enzyme StuI, which has a recognition site only in URA3 coding sequence but no other region of the construct. Then the yeast competent cells were prepared as follows: a single colony of YM4271 strain (MATa, ura3-52, his3-200, ade2-101, ade5, lys2-801, leu2-3, 112, trp1-901, tyr1-501, gal4d, gal8d, ade5::hisg) was inoculated into 2ml YPD (YPD Broth, Sigma- Aldrich) liquid medium and cultured by shaking at 30 C for 16 hours; the yeast culture was diluted 1:2000 in 50ml YPD liquid medium and cultured by shaking at 30 C for 20 hours to an OD 600 value of 0.75; the cells were harvested by centrifugation at 700 RCF for 5 minutes at room temperature; the medium was decanted and the cells were resuspended and washed in 5ml sterilized water followed by a centrifugation at 700 RCF for 5 minutes at room temperature; the supernatant was discarded and the cells were resuspended in 1ml freshly made TE/LiAc solution and then another 5-minute centrifugation at 700 RCF was performed; supernatant was discarded and the cells were re-suspended in 400μl TE/LiAc solution; 40μl denatured salmon sperm DNA (prepared 39

57 by boiling a tube of 2mg/ml salmon sperm DNA for 10 minutes and then incubating on ice) was added in the solution. Then for each transformation reaction, 100μl yeast competent cells prepared as above were added into the linearized DNA bait:: reporter construct, and 600μl freshly made TE/LiAc/PEG solution was added and mixed. The mixture was incubated at least 30 minutes at 30 C followed by exact 20-minute incubation at 42 C. Then the transformed competent cells were precipitated by centrifugation for 5 seconds at full speed in a microfuge, and 300μl sterile water was used to suspend the cells. The cells were spread onto URA - selective medium, containing 6.7g/L yeast nitrogen base without amino acids (Sigma-Aldrich, Y0626), 1.92g/L yeast synthetic drop-out medium (Sigma-Aldrich, supplement without uracil Y1501), 20g/L Bacteriological agar (A5306), 40ml/L glucose (50% w/v stock). After 3 days incubation, colony PCR was performed to select the positive cells having the construct. First, 6 single colonies were streaked onto YPD plate and cultured at 30 C for 2 days (the colonies are about 2mm in diameter). A quarter of a colony was re-suspended in 15μl Zymolase (Zymo Research Corporation, concentration is 2mg/ml) in 0.1 M Sodium phosphate buffer (ph 6.8). Then the mixture was incubated at 37 C for 30 minutes followed by 95 C for 10 minutes. Cells were spin down and the lysate was diluted by adding 10μl upper liquid to 56μl water, and 3μl was used as template for PCR reaction. Two sets of primers were used check the insertion and the insert numbers. N-3788: TGCCACCTGGGTAATAACTCG/N-3790: TATTCTTAACCCAACTGCACAGAAC were used to check the integration of the plasmid into yeast genome. N-3789: CATTTGCTTTTGTTCCACTACTTTT/N-3790: TATTCTTAACCCAACTGCACAGAAC were used to check the insertion number Yeast one hybrid screening 40

58 The screening procedure was performed by our collaborator, Jose Pruneda-Paz lab in UCSD. Bait strains were generated by homologous recombination of placzi (Clontech) bait vectors in the yeast YM4271 according to manufacturer s protocol (Clontech). The transformation of the TF library to each yeast strain was performed in a 384-well format. The β-galactosidase (β-gal) activity was determined as described by the manufacturer (Clontech) but with modifications that allowed the assay to be performed in 384-well plates. Briefly, transformed yeast were transferred to 384-deep well plates, and grown for 24 hours at 30 C in 400µl of SD medium lacking tryptophan. After incubation, 100 µl of the culture were transferred to a new 384-deep well plate containing 400µl of YPD and grown for 3 hours at 30 C. A 150µl aliquot of this short-term culture was used to determine the OD600 using a 384-well plate reader (Spectramax Plus, Molecular Devices). Yeast cells from a second aliquot of 200µl were washed with Z buffer (10 mm KCl, 1 mm MgSO 4 in phosphate buffer ph 7.0), re-suspended in 30µl of the same buffer, and lysed by performing four freeze/thaw cycles. The enzymatic reaction was started by the addition of 170µl of Z buffer/ 120µg ONPG (2- Nitrophenyl-β-D-galactopyranoside) (Sigma) to the lysate, and was incubated at 30 C between hours. Finally, the enzymatic reaction was stopped by the addition of 80µl 1M Na 2 CO 3, cleared by centrifugation and 150µl of the supernatant used to determine the OD420 using a 96-well plate reader (Spectramax plus, Molecular Devices). β-gal activities for each TF were calculated and normalized to the average value obtained with the pexp-ad controls. 2.5 Bioinformatics data analysis The spatiotemporal transcriptome was generated by a previous post-doc, Jeffrey Choonwei Wee in our lab. The dataset and the experimental information are available 41

59 (Geng et al., 2013). All data analyses were performed using the R software package ( and packages provided through Bioconductor ( GCRMA was used for global normalization (Irizarry et al., 2006). Probe-sets that are annotated by TAIR to hybridize to multiple loci in the Arabidopsis genome were removed from further analysis based on the affy_ath1_array_elements table. The original normalized expression values of different samples were used for PCA analysis in MEV (Saeed et al., 2003). Differentially expressed genes were determined using LIMMA and EDGE packages in R (Smyth, 2005; Storey et al., 2005; Leek et al., 2006). P-values were corrected for multiple testing using the Benjamini-Hochberg method and probe-sets were considered significantly differentially expressed if the P-value and expression differed by at least 2-fold between contrasting sample types. A 10% FDR threshold was used in EDGE combined with a 2-fold change cut-off. The final list of NaCl regulated probe-sets were attained by combining the lists from LIMMA and EDGE and subtracting age regulated probe-sets. A soft clustering algorithm, fuzzy C-means was used to partition the significant differentially expressed genes based on their spatiotemporal expression patterns (Hathaway et al., 1996). A probability cutoff of 0.5 was chosen as this led to stringent clustering and assignment of each gene to only one cluster. Due to this stringency some genes were not included in any cluster as they did not correlate strongly with any of the predominant patterns found. Different C values (number of clusters) were tested from 15 to 50. Enrichment of GO terms was used to evaluate each C value (Ashburner et al., 2000; Brady et al., 2007). Partitioning the data into 25 clusters led to the largest number of enriched GO categories with 3897 genes being clustered in total. GO category enrichment was performed using the ChipEnrich JAVA application (Brady et al., 2007). The 25 clusters were subject to CREs (cis-regulatory elements) enrichment analysis using online 42

60 version FIRE (Elemento et al., 2007) for both known and potential unknown elements. FIRE uses information from the assignment of genes into clusters in combination with the sequence of the upstream promoter region to identify enriched seed sequences. These sequences are then optimized to generate a position-weight matrix, which describes the best scoring pcre motif. In addition, orientation and positional bias of the pcre can be evaluated by FIRE. The salt clusters were also subjected to CREs enrichment with the website-based Athena algorithm ( in which almost all the characterized DNA binding proteins are collected. Hormone treatment data sets were generated by the AtGenExpress consortium and described previously (Nemhauser et al., 2006). Original.cel files were downloaded from Probe hybridization values were extracted and normalized using GCRMA (Irizarry et al., 2006). Pairwise comparisons were made between hormone treated and control samples at each time-point using LIMMA. Genes were ranked based on the B-statistic of differential expression and the top 400 genes with the highest values were chosen for further analysis. K-mean clustering was used to group genes into 7 submodules using the city-block distance metric. Gene membership was compared between the salt-stress clusters and the hormone sub-modules and the significance of shared membership calculated using the hypergeometric probability. P-values were corrected for multiple testing by calculating their Q-values (Storey and Tibshirani, 2003). A network diagram illustrating the inferred secondary signaling network was visualized using Cytoscape (Shannon et al., 2003) and are available for viewing at Programs used such as Python, Django, and Cytoscape were run in the iplant atmosphere. This bioinformatics analysis of large amount of datasets connection was aided by the computer expert, Xueliang Wei in our lab. 43

61 2.6 Live imaging Seedlings were grown and imaged using a custom macroscopic imaging system as previously described (Duan et al., 2013). Specifically, samples were manipulated using a circular platform with six square tissue-culture plate holders, which is controlled by an automated Theta/360 degree rotary stage and MFC-2000 controller (Applied Scientific Instrumentation); samples were backlit using an infra-red LED panel, and images were captured using a digital monochrome camera (CoolSnap) fitted with an NF Micro-Nikor 60mm lens (Nikon) and infra-red filter. Micro-Manager Software (Vale Lab, UCSF) was used to control the stage and automate image acquisition. Images were taken every 15 minutes for up to 24 hours. Sequential images were collated as a stack for further analysis using imagej (Abramoff et al., 2004). The StackReg plug-in was used to align the stack of image slices before root growth quantification. The Growth rates were quantified using a semi-automated image analysis algorithm written as a macro for and is available through our lab s website ( The macro processes the image stack to enhance the contrast of edges. The user then draws a line in the last frame of the series along the midline of the root from the tip to the base. The algorithm uses the midline as a landmark to aid in identifying the position of the root tip in each frame and translates this position back onto the midline. A table is generated listing the distance between root tip positions between subsequent frames of the time-lapse series. 2.7 Confocal microscopic analysis For confocal microscope imaging, roots were mounted in an FM4-64 solution (Invitrogen), and imaged using a Leica SP5 point-scanning confocal microscope. The imaging settings are 488 nm excitation and nm emission for GFP, 514 nm 44

62 excitation and nm emission for YFP and 488 nm excitation and >585 nm emission for FM4-64. For quantification of RGA-GFP, confocal images were taken as a Z-stack with the same distance from the middle layer to the surface layer of root tips. GFP signals were measured using image J as mean gray value for a specific area for each stacked image, and the mean value of background of the same area was deducted from each measurement. Ten roots were analyzed for each condition and the average was calculated. Significance was calculated using student s t-test. 2.8 GUS staining Seedlings were mounted in a modified staining solution of 1mM X-Glc, 0.5% (v/v) dimethyl formamide, 0.5% (v/v) Triton X-100, 1mM EDTA (ph 8), 0.5mM potassium ferricyanide, 0.5 mm potassium ferrocyanide, and 100 mm phosphate buffer (ph 7). The samples were incubated in 37 C, and the time varied for different samples. For the synthetic promoters of ABRE, 30 minutes was used for the staining. For the synthetic promoters of MYC binding motif, AACCACT, and ATATAAT, 15 minutes was enough for the staining signals appearance. And for the rest, wbox and telobox, longer time was needed, which were 2 hours and about 12 hours, respectively. After staining, the samples of roots were cleared with Hoyer's solution, which is chloral hydrate:water:glycerol in proportions 8:2:1, LUC analysis D-luciferin sodium salt (Gold Biotechnology, LUCNA-100-SPO) was dissolved in 1x PBS to get a stock concentration of 200mM. Seedlings growing on square petri dish were 45

63 sprayed with 1mM luciferin in 0.01% Triton X-100 solution and kept for 1 hour before imaging. The imaging system (Bioimaging Solutions, San Diego, CA) was designed by a post-doc, Ruben Rellan, in our lab. It is based on two PIXISXB 2048 x 2048 pixels CCD cameras (Princeton Instruments, Trenton, NJ), and the whole system is mounted inside a light tight enclosure. Custom software ((Bioimaging Solutions) was used for the system to capture the luminescent images Gene expression Total RNA was purified using the RNeasy Plant Mini Kit (Qiagen GmbH) according to the manufacturer s instructions. Quality and quantity of each RNA sample were analyzed using a Nanodrop spectrophotometer. cdna was prepared using the iscript advanced cdna synthesis kit (Bio-Rad) from 600 ng total RNA. The high throughput Q-RT PCR was performed on a Fluidigm BioMark Expression Chip using EvaGreen (Bio-Rad) as the fluorescence probe according to the Fluidigm Advanced Development Protocol #37. Two control genes AT4G37830 (cytochrome c-oxidase) and AT3G07480 (electron carrier/iron-sulfur cluster binding) were used for data normalization. Expression values were calculated using the 2 -ΔΔCT method. For statistical test, two to three technical replicates and three biological replicates were generated per sample type. Student s t-test and 2-way ANOVA were used to test for the statistical significance in aba-sail experiment and transactiviation lines experiment, respectively. And the results were visualized as a heatmap generated using ΔCT value in TM4-MEV (Saeed et al., 2003). For testing dynamic gene expression under salt stress in aba2-sail mutants and abi1-1 transactivation lines, the primer sequences and information of the 94 test genes are listed in Table 3. Whole roots were tested for aba2-sail mutants experiment and root tips 46

64 (meristem and elongation zones) from fifteen to twenty seedlings were tested for transactivation lines experiment. 47

65 Table 3. Accession numbers of analyzed genes and primers sequences used during the Real-time quantitative PCR analysis. AGI forward reverse Salt hormone cluster sub-module AT5G19510 AGGAGAGGGAGGCTGCTAAG CACGAGGTCATCAACAATCG 1 aba_2 AT5G16070 CTGTGGTGGAGAAGCTGTGA CCCTTGAACGGTTTTCTTGA 1 aba_2 AT3G59540 CCAGGTTTGAGTGTGCAAGA CCAAACTGTTGGGGCCATA 1 aba_2 AT3G57290 GAACCTAATCCGCACCTCAA ATTCCAGTCAACCGAAGGTG 1 aba_2 AT4G26230 GAGGAGGTGGTGACCAGAGA GTCCTTCAGCAGGGATTTCA 1 aba_2 AT3G16810 GACCCCCTTGTGAGAAGACA ACGACCGACATTTTCCAGAC 1 aba_2 AT3G18130 GTGGTGGGAAAGATGGTGTT TTTTCCGGAGATTTCGTGAC 1 aba_2 AT3G09700 CTGTTGCTGCTGCTGCTTAT CCTCCTGCATCTGGATGATT 1 aba_2 AT2G20450 CTACGGCGAGGATTATGGAA TTGGCCAACATGATCTTGAA 1 aba_2 AT1G54690 GTCTTCAATTTCCCGTTGGA TCATCGTTCCTCACTGCAAG 1 aba_2 AT2G34260 CTCCAAATTCCGTTGATGCT CCAGATGCATTCCCAGAGTT 1 aba_2 AT5G57050 GATCACAAACCGGATAGGGA CCATCGCGTTCTTCTTATGC 6 aba_1 AT5G06760 AGGCGGAGAAGATGAAGACA GTTGTCCGACCAGTTCCAGT 6 aba_1 AT4G26080 TCTCAGGTAGCGAACTATTGTAG TGGTCAACGGATAATGGAAGTG 6 aba_1 AT1G49450 AGCGGTGGAGCTGATAAGAA CGCCACACTTTAACGGAGTT 6 aba_1 AT2G46680 GACCGGAGATGGAGATGAAA CTCGGTTCATGCGATGTAGA 6 aba_1 AT5G04250 CCGATACCAGAAGGCAAGAA GGGGTCCCTCATAGGAATGT 24 aba_7 AT3G23920 GACTCTGTCACTATTCCTCTGC AACTTCCATGTCCCTTCTTGCTC 24 aba_7 AT3G06500 TTGCAAACATGCCTCTCAAG CGCTATGGTCCAAGTCTGGT 24 aba_7 AT3G03170 GACCGTCGCAGTTCTTCTTC TTCCGGTTTAGGTTCGAGTG 24 aba_7 AT1G30620 TGGTAAGGCCAAGAAGATGG GATTCCAACTTTGCGTGGTT 24 aba_7 AT1G69260 GGAAACTAAAGCGTCCAGTGAC ATGGCAAACACACATGATCC 24 aba_1 AT1G60190 CGGTGGTGGATTAAGGCTAA CCGCCAGTGATTATCAGGTT 24 aba_1 AT1G62570 GCTAGCTCCCGGACTTTCTT GTGGGAATCCACACAGGTCT 24 aba_1 AT1G51140 AGGACCCAATCTGGAGGTCT TTGTTTGCGTGTCCATGTTT 6 aba_7 AT5G57610 TGGTGTGATCCAGAGTGGAA GTTCCAAACGCAAACCAAGT 6 aba_7 AT1G13740 CTGTGGTGTTGATTCCGATG ACGTAGCCATCCCTTTTCCT 6 aba_7 AT1G72770 ATTGGTAGAGCCGTTGTTGG AATGTCACTTCCGGTTCTGG 6 aba_7 AT5G59845 TGAAATTCCCGGCTGTAAAA CGGTAACAAGGGCATTCATC 5 aba_5 AT3G53960 TCTTTTGAGACGTGCCACAG GACCGACAAGGGTAAAAGCA 5 aba_5 AT4G35480 TTCAGTCGGAGACGGAGATT GGAATAATGGCGGAAGTGAA 5 aba_5 AT3G25620 TGGGGATTCTTCCCTCTCTT CGTCCATCAAGATTGCTCCT 5 aba_5 AT1G47960 CTAAAACGGGCTTTGGATGA TGAAACAACACAAGCCTTCG 5 aba_5 AT5G14130 CCCAACCCAGATGCTGTAGT CAGAGCAGTCCCTTCGAATC 3 aba_5 AT5G10230 CAGGAGCAAAGCACAAATCA ATGTCCGAGAAGAGCGAGAA 3 aba_5 AT5G08250 TGCTGGGAGAGACACTTCCT TCCCCCAAATTGTTTCCATA 3 aba_5 AT5G05390 GCGTTTTCACAACGGATTTT TCAACCAAACCCCTGGATTA 3 aba_5 48

66 AT2G33380 AGCACCTATGACACCGAAGG AGAATTGGCCCTCTCTTTGG 13 aba_1 AT5G52300 GGAGGTGGAAGTGGAGTGAA ATCCGAAAACCCCATAGTCC 6 aba_4 AT1G16850 ATCGTGGACAGTCCCACTTC CTCCCCCAAAACTGTGTCAT AT3G02480 CCTTGCAACAGACTGGACAA AATGGACGCAAGGAAACAAC AT1G04560 CTGCTGGAGCTTCTTCCATC ATGAGCTTGGCCTGTAGCAT 6 aba_4 AT2G42580 ATGGTGGATGTGGAGGAGAG AGCAGAGCCAAACTCCAAAA 21 br_4 AT5G60920 GAGAAAGGTTGGGCTTTTCC CCGTAACCCTAAGCAAACCA 19 br_4 AT1G12500 CCGGAGTCTCGGTTTTGATA CCCTCTCAAAACGAACCAAA 19 br_4 AT1G61100 ACGGAACAGCTCAGGAGAAA GGCTGATCTCTCTGCGATTC 19 br_4 AT2G22125 GCCTTGGATGCATTGTTTCT TGCCGAGTGTTATCTTGCAG AT3G23750 CCTCAACAAAGACCCGACAT GCCAAAGCCAAAATGAAAGA AT5G07110 CTCTACGCGTCCTCTTGTCC GGTCATGAACATCACCACCA 21 br_4 AT1G68410 TAGCTCCAAGCGAGGGAATA GAACCCGATTCTCAGATCCA 19 br_4 AT2G23130 GGCTCCAGCTCTGACAAAAC GCAGCTCCCACCATTTGTAT 19 br_4 AT5G16590 CGACAACATCTCATGGCAAC TAACTGCTGATGCGTTGGAG 19 br_4 AT1G03457 AACGGCACATCGATACCTTC TTACCGCTTAATTGGCAACC 19 br_4 AT1G24170 TGTTTGCGGGTAATGTTGAA CTCCGACAGAGAAGCAAACC 19 br_4 AT3G28200 CGTGATTCAAGAACGTCGAA CACTGCGAATCTCGGGTTAT 21 br_4 AT5G04310 CGGAAGGTGACGTTATGGTT TCAATGTGAAAAGCCCATCA 19 ga_5 AT5G03760 TCGGTTTACTCGAAGGAGGA GATTGCTTGTGCGAAAAGGT 19 ga_5 AT1G22330 CAGCTCCAACAGGCACAATA GCGTGTGATTCAGTGGAAGA 19 ga_5 AT3G07010 GCCAAGATGTAGGCATGGAT CAGCTCCAGAACGTGTGAAA 19 ga_5 AT5G17640 TCAGAGACTCCGTGTGTTGC AAAATGGCTGCATCTTCCAC 21 ga_4 AT1G76240 ACTGTTTTGGTTTCCGTTGC TTTCACGGATCGAATTCTCC 21 ga_4 AT5G15350 GTGTTCGACAGAAACCAGCA TGAATTGAGCGAGACCTGTG 21 br_4&ga_4 AT1G76670 CCAATCCAAGCCATTTCACT CGTTGACTTTGCGTTCCTCT 21 br_4&ga_4 AT5G60670 GCGGCTCTAGTCATCAAAGC CCATCAACAGTGCAACCAAC 1 zeatin_3 AT3G61100 AGGAGATGCAAATGCGAGAT AAGCTCGGGTGATAGGGTCT 1 zeatin_3 AT3G56070 TGGCTAACTCTGGTCCCAAC TCAAGATCAACCCACCCAAT 1 zeatin_3 AT3G49080 CATTGCACGTGTTTGGATTC CTGGTTCCCAGTTTTGCAGT 1 zeatin_3 AT1G06720 AGGCAATCGAAAGAAAGCAA GTGATGTTGAAGCAGCTGGA 1 zeatin_3 AT1G56110 TAAGAACAAGGGCCGTATCG CATTGCCACCACCTCTTCTT 1 zeatin_3 AT1G15250 CGCTCTGTGTGAGATGTGGT GCAAAACGCAAACTGAACAA 1 zeatin_3 AT1G70310 AACGCTGCTGAAGGAACCTA CTCCACTCGGGTAAGTTGGA 1 zeatin_3 AT2G40660 GAAGCCTGCTGAACCAGAAC GCGGTTCGTTAGATCCTCAG 1 zeatin_3 AT2G19720 TTCCAAGTCCATGATCCACA GAACCTGACCACCCACATTC 1 zeatin_3 AT4G30800 AACATTGGCCTTGGTTTCAA GCCCAATGGTAACACGATCT 1 AT3G57490 TTCTCGCGAAATTGTCTGTG GCCTTGACGAAGTTTCCAAG 1 AT2G40010 ACCTCCTTCTTCCAGGTGCT ATTGCAAGAGAAAGCGCAGT 1 AT1G64880 TGCAAAAGCAAAAGCTGAGA AGGCCTTCAACACTGCCTTA 1 AT3G10610 CAATCATCCCATCGAAAAGG AGCAGGTGCCATAGCTTGTT 1 49

67 AT1G07070 GTCAGGGGAACCATACTTGG TGATGTGAATTTGGCTCGAA 1 AT5G53070 CTGGCGAATGCTAATTTGGT CCTTGTGGCAAAGGAATTGT 1 AT4G25890 ACCTACGAGCTCCAGCGTAA TCCGAAGTCTCCTTCTTCCTC 1 AT2G32060 AGTGATCCCAGAGGACATGG CCAACAACCTTCCTTGCATT 1 AT3G44750 GCCACAAGGCTATTCTGAGG CTTTCGCCTTCTTTGCTGAC 1 AT5G56580 AAAACCCGCCAAGCTTTTAT AACAGGTGGTTCCAGAGTGC 1 AT1G08090 CGCCGAGTACTTCTTTGACAGGT ATAGAGAAGAGCACCATAGCCAC AT5G66400 AAGATCAAGGAGAAGTTGCCAGG GTAAACAACACACATCGCAGGACG 6 AT2G36270 ACAGCAAATGGGAATGGTTGG AACTCCGCCAATGCATGTTT AT3G61890 AGCAATCTCTGGTCTCTGAGC TCAAGCAACTATCATCAGCTTTC AT4G34160 CCTCTCTGTAATCTCCGATTCAA AAAGGGTTTGCATCAATCACG 21 AT1G21410 TCTAGCAGACTGGTGTGTCC GTTAAGTTCAGACAGCCGCTC AT4G33950 GGTTGCAGATGTTTGGTCTTG ACTGAGTGGTCATCGTGTTATC 20 AT5G52310 GCACCCAGAAGAAGTTGAACAT GAATAATTTCCTCCGATGCTGG 6 AT5G08640 TAGCTTTAGGTGTACCGGCTC TTCCGGTAAAGGTCCAACAATC 16 AT4G14550 CTAATCAGAAGAGCGGCGAAG AGCATCCAGTCACCATCTTTG 7 AT4G37830 GCGATTGTACGTTCAGCTCTTTC GTGCTCTTTGTTGTGCTTCACC AT3G07480 CTCTTCAGAAACTCTCCTCTCAA ATTCCTCTGCGATCTGAACCTC 50

68 2.11 Genetic analysis To selectively express abi1-1 in different tissue types, various enhancer trap lines were crossed into homozygous UAS:: abi1-1 plants. Wild-type plants of C24 were crossed with UAS:: abi1-1 plants to generate the control genotype. Gene expression analysis was performed using the F1 seeds. 51

69 Chapter 3 Results and discussions I 52

70 3.1 Abstract Much work has been focusing on understanding the function of hormone signaling during environmental stress response in plants. However, questions remain, such as how the biosynthesis and signaling of different hormones are integrated together during stress response in a spatiotemporal manner, and what downstream transcriptional modules are controlled by hormones during the response to environmental stress. In order to answer these questions, we first generated a spatiotemporal transcriptional map of salt stress in Arabidopsis roots, covering 4 core cell types and 6 time points for salt treatment. Compared with a previous study showing tissue-specific responses at 1 hour to high salinity, this map provided higher temporal resolution, giving a more dynamic view of how different cell types respond to salt stress at different time periods of salt treatment. Based on this spatiotemporal map, the transcriptional changes of key components in hormone biosynthesis and signaling are identified, suggesting that these hormones function in specific cell types and at particular stages during acclimation to high salinity. A bioinformatics method was also developed to systematically de-convolve the hormone crosstalk network with salt stress, identifying some salt stress response sub-modules controlled by hormone signaling. A good portion of these modules were validated using high throughput q-rt PCR. Taken together, this study is valuable in demonstrating the transcriptional response to salt stress at temporal and spatial scales, and systematically revealing the hormone crosstalk with salt stress in Arabidopsis roots. 53

71 3.2 Introduction As described above, multicellular organisms are composed of complex assortments of tissues and cell types that must coordinate biological activities to enable normal development and survival from environmental stress. The root of Arabidopsis has become an ideal organ system to study the response of plants to changes in the environment. The Arabidopsis root is essentially composed of concentric layers of tissue types with different identities and biological functions, surrounding a central core of stele tissue where the vasculature is housed (Benfey and Scheres 2000). FACS has proven useful for exploring the cell type specific response of root tissues to abiotic stresses, such as nitrogen content (Gifford et al., 2008), high salinity (Dinneny et al., 2008) and nutrient deprivation conditions (Dinneny et al., 2008). High salinity stress, which we used as a model in our study, is a very important contaminant in agriculture that can cause the growth retardation through secondary stresses such as ionic stress and osmotic stress (Xiong 2002). Dinneny et al. (2008) illustrated that the cortex is the most responsive cell type in Arabidopsis root by profiling and comparing transcriptomes of different cell layers using FACS. However, to be noted, this work is restricted to a one-hour treatment by salt stress. Recent work examining osmotic stress responses in the shoot have highlighted the dynamic nature of the acclimation processes (Skirycz et al., 2010, 2011). Upon treatment of seedlings with mannitol, a water-stress stimulant leaves exhibit temporally dynamic changes in gene expression and growth. Also, under a temporal salt stress condition, lateral root of Arabidopsis experiences a dormant phase in growth rate and dynamic gene expressions (Duan et al., 2013). Both of the changes have been shown related to the spatiotemporally regulated fluctuations in hormone signaling such as ABA signaling. For example, Duan et al. (2013) has shown that the early repression of growth rate of lateral 54

72 root is dependent on an increase of ABA and endodermis specific ABA signaling, while the recovery of growth from the dormant phase is due to the decrease of ABA. Here I described the generation and analysis of a spatiotemporal transcriptional map for the salt-stress response in roots. Using FACS-mediated tissue-specific expression profiling, we generated time-course data sets for each of the four major tissue layers. These data provided a highly resolved resource that enables the use of spatial and temporal information to generate hypotheses regarding the regulation of these expression patterns. Environmental responses in plants are regulated, in part, through a complex secondary signaling network enacted by changes in hormone biosynthesis (Dinneny et al., 2008). I also developed a novel bioinformatics method that utilizes publically available hormone-response data sets to de-convolve the secondary signaling network involved in controlling discrete transcriptional programs regulated during the salt response. Finally, tissue-specific strategies for the manipulation of cytokinin or ABA signaling revealed an important role for inner tissue layers in controlling ABA-mediated salt stress transcriptional responses. Together, our analysis provides a highly resolved understanding of the regulatory networks that generate the complex expression patterns we observe during a stress response. 55

73 3.3 Results A global spatiotemporal transcriptional map of the salt stress response in Arabidopsis root In the previous study, a high resolution spatial transcriptional map in response to salt was generated using 1 hour treatment with 140mM NaCl, including six root tissue layers (Dinneny et al., 2008). This data set reveals the extensive tissue-specificity of the response. However, live imaging data generated by Xie Fei in our lab indicated that the response of root to salt is temporally dynamic and additional transcriptional programs can be elicited at the later time points. Moreover, to facilitate how different hormones systematically integrate into salt stress response in a spatiotemporal manner, the spatiotemporal transcriptional map was decided to be generated of the model plant organ, Arabidopsis roots, in salt stress response. The microarray data set was generated by the previous post-doc, Dr. Jeffrey Choonwei Wee. The experimental information and the raw data are available (Geng et al., 2013). To briefly introduce this data set, four different GFP-based reporter lines, ProWER::erGFP, ProCOR::erGFP, ProSCR::erGFP, and ProWOL::erGFP were used for FACS isolation of epidermis (including lateral root cap), cortex, endodermis, and stele cells, respectively (Figure 2). The time series for salt treatment were 1, 3, 8, 20, 32 and 48 hours to trace the dynamic gene expression changes during salt stress response. Controls of standard condition were performed for 1 and 48 hours to test the developmental (age-dependent) effect. To first evaluate the quality of this data set, I tested using the previously published developmental genes and they showed consistent expression patterns in this data set with the reports (Figure 3). 56

74 To have a general understanding of the relationships between the data points of the spatiotemporal map, I did the Principal Component Analysis (PCA) on the data set to find the associations between the various cell types and time points. Based on the first two principal components (58% of variation captured), it was found that the relative relationship between the transcriptional profiles of each cell type, with respect to the other cell types, was maintained throughout the salt-stress response (Figure 4). Specifically, the first principal component (38% of variation captured) that usually captures only gene expression levels separated the cell types, suggesting a cell type specific transcriptional responses caused by cell identity. The second principal component (20% of variation captured) captured a common featured transcriptional response to salt among different cell types (Figure 4). Since the data points coming from the two standard conditions (M1 and M48) were dramatically separated from the other data points of salt treatment, it suggested the salt response happened in all cell types. However, the temporal salt response could be different among cell types, because the early response in stele showed a different pattern from the ones in other cell types (Figure 4). To identify genes whose expression changed dynamically during the salt response in each cell type, stepwise comparisons were performed between subsequent time points (e.g. Standard 1 hour vs. NaCl 1 hour, NaCl 1 hour vs. NaCl 3 hours, etc.) for each cell type using LIMMA and 4344 significant differentially expressed genes were identified (gene list available in Geng et al., 2013). Also, another list of significant genes were obtained using a time-course specific statistical package, EDGE, which models temporal data as splines to take advantage of the non-independent nature of samples collected over and 2,996 significant genes were identified (gene list available in Geng et al., 2013). Based on these two methods, a combined list of 5,990 genes was considered as significantly salt regulated. Similar to the previous findings (Dinneny et al., 2008), the 57

75 cortex cell layer was the most transcriptionally active cell layer throughout the time course (Figure 5). To better reveal the dominant patterns of transcriptional response during salt stress, the above salt-regulated genes were organized into 25 co-expressed clusters using the mfuzzy C-means algorithm (Figure 6) (Hathaway et al., 1996). Several modes of transcriptional regulation can be distinguished by the level of cell-type specificity and temporal dynamic nature of the transcriptional response. For example, cluster 6 showed highly coordinated activation in gene expression across all cell types early in the salt response time course and was enriched for ABA responsive genes (P-value < 1E-8, Figure 7) such as RD29A (Figure 6). Cluster 3 is an example of tissue specificity and dynamics, showing peak expression in the endodermis layer between 3 to 8 hours after salt treatment, such as LAC13 (Figure 6). This cluster was enriched for biological functions associated with Casparian strip formation including suberin biosynthesis (P-value < 1E-4, Figure 7) and laccase activity (P-value < 1E-6, Figure 7), which is necessary for lignin biosynthesi. And this strengthened structure can be confirmed with CAPS1-GFP reporter treated by salt (Figure 7). 58

76 Figure 2. Generation of spatiotemporal transcriptional map. (A) Confocal images of ProWER::erGFP (a), ProCOR::erGFP (b), ProSCR::erGFP (c), and ProWOL::erGFP, which were used in the FACS isolation of specific cell layers in this study. (B) Five day seedlings were treated with 140mM NaCl for 1, 3, 8, 20, 32 and 48 hours; seedlings were also transferred to standard condition for 1 and 48 hours as controls for the developmental effect. (C) Workflow for the FACS (Fluorescence activate cell sorting) experiment. Salt treated roots were collected and protoplasted, and the obtained cells were conducted with FACS and the GFP positive cells were collected into tubes, which were used for the RNA extraction and microarray. *Data generated by Dr. Jeffrey Choonwei Wee. 59

77 Figure 3. Expression of developmental genes in the spatiotemporal map under salt stress. The developmental genes from the literatures that are tissue-specifically expressed were used to check the quality of the spatiotemporal map. 60

78 Figure 4. Principal component analysis of the different sample types composing the spatio-temporal map of the salt response. The first principle component is 38%, and the second principle component is 20%. M, standard conditions; S, salt stress conditions. Numbers represent the time treated by 140mM NaCl. Blue, epidermis; purple, cortex; orange, stele; green, endodermis. 61

79 Figure 5. Number of genes that showed differential expression in each cell layer at different time points after salt treatment. The graph was generated using the significant differentially expressed genes for each separate cell type, including epidermis (A), cortex (B), endodermis (C), and stele (D). The method used is stepwise comparison using LIMMA. X axis, the pairwise comparison; Y axis, number of significant genes. Red, up-regulation; blue, down-regulation. 62

80 Figure 6. Spatiotemporal expression patterns observed during the salt response. (A) Centroid profiles for each of the 25 gene clusters identified in the spatiotemporal data set. (B) efp representation of RD29A expression from cluster 6. (C) efp representation LAC13 expression from cluster 3. *web-based efp browser ( appearance was generated by Xueliang Wei. 63

81 Figure 7. Biological processes regulated in spatiotemporal salt stress response. (A) Select enriched GO categories for each cluster shown. The AGIs in the 25 clusters were used for the GOs analysis using chipenrich software. Cut off of p value is (B) An example showing the structural change in casparian strip under salt stress, which is based on cluster 3. The structure of casparian strip is labeled by CASP1: :CASP1:GFP (Roppolo et al., 2011). Images of longitudinal sectioning and reslice projected cross sections were shown. 64

82 3.3.2 Different strategies were used to adapt salt stress at early and late stages To gain an understanding of how the initial salt response differs from the late recovery response, the analysis on genes regulated by salt at 1-hour or 48-hours after treatment was conducted. 3,034 genes regulated at 1 hour and 6,054 genes regulated at 48 hours were identified by LIMMA (Geng et al., 2013). Genes activated during the initial response tended to be annotated as responsive to ABA and a broad range of stress conditions (Figure 8), suggesting that the stresses elicited early on by high salinity are common targets of ABA signaling and other environmental conditions. Genes that are immediately down-regulated were enriched in cell wall biosynthesis (COR-enriched), which correlated well with the immediate suppression of growth and radial cell expansion observed after salt treatment (Figure 8, and Dinneny et al., 2008). These early observations are consistent with the previous study (Dinneny et al., 2008). At 48 hours, a massive increase in the expression of genes associated with protein translation including large and small subunits of the cytosolic ribosome (Figure 8) were observed, and this late up-regulation of protein translation may be associated with long-term acclimation and the resumption of growth in the root (Geng et al., 2013). Late down-regulated biological functions include water channel activity (Figure 8), which may prevent water loss due to growth in a low water-potential environment (Chrispeels and Agre, 1994). 65

83 Figure 8. Biological processes involved in early and late stages of salt stress responses in different cell types. (A) GO categories for the activated genes in the early and late stages. (B) GO categories for the repressed genes in the early and late stages. Genes showing significant expressions in the early and late stages were used for the GO analysis. 66

84 3.3.3 A cluster-comparison method identifies targets mediating hormone signaling in salt stress response Plant hormones are important intermediary signals controlling growth downstream of environmental stimuli (Jibran et al., 2013; Duan et al., 2013). In order to systematically explore the transcriptional targets of hormone signaling in salt stress response, a bioinformatics method was developed that allowed us to identify the known hormonesignaling pathways which might contribute to the various expression patterns observed in this data set. To set up the connection between salt stress and hormones, I downloaded the public AtGenExpress hormone treatment data sets, utilized LIMMA to identify genes differentially expressed and performed k-means clustering to segregate these genes into sub-modules based on their temporal response profile. Genes in these sub-modules were then analyzed for their membership in the set of 25 salt-responsive clusters that was identified in the spatiotemporal map (Figure 9). A network diagram including 16 hormone sub-modules and 13 salt responsive clusters was generated with a significant overlap (Q-value < 0.01) with the assistant of the computational expert, Xueliang Wei (Figure 9). ABA signaling was the most informative, since it was associated with the largest number of salt-responsive clusters (Figure 9). Meanwhile, this network can give us a prediction of hormone crosstalk in the salt stress response. For example, BR and GA pathways, which have recently been shown to act synergistically to promote growth through direct protein-protein interaction of signaling pathway members (Bai et al., 2012, Gallego-Bartolome et al., 2012), target the same set of salt-regulated clusters (19 and 21) (Figure 9). Also, the crosstalk between ABA and cytokinin can be mediated through cluster 1 (Figure 9), which was activated at late stage. 67

85 Figure 9. Analysis of the hormone secondary signaling network regulating salt-dependent transcriptional programs. (A) Comparative methodology for generating the network between salt clusters and hormones clusters. The publicly available data sets with different hormone treatment were analyzed by LIMMA, and the significant genes were clustered based on the patterns showing up or down regulation at each time point. The overlapped genes were analyzed between the hormone clusters and salt clusters. (B) The expression patterns of hormone clusters. Centroid values were used to generate the heatmap. (C) ABA activating sub-module. (D) ACC sub-module. (E) ABA suppressing sub-module. (F) GA and BR sub-module. (G) JA sub-module. * The method was developed by me, and the web-based network (shown online) was generated by Xueliang Wei. The thickness represents the significance of correlation. The symbols of arrows and blocks represent activating or repressing roles of the hormones in regulating the salt cluster 68

86 3.3.4 Spatiotemporal understanding of hormone biosynthesis and signaling pathway ---ABA as an example From the network, it can be predicted when and where a hormone was induced to be synthesized and integrated in the salt stress response, as well as the targets for response. These predictions would be very important for the biological output, such as salt tolerance. Since ABA was well studied previously, I took it as an example for verification purpose. Our inferred secondary-signaling network suggested that ABA has the most diverse role in the regulation of transcriptional programs during salt stress (Figure 9C). Six saltresponsive clusters showed significant gene membership overlap with signaling submodules induced (clusters 3, 5, 6, 13, and 24) or repressed (cluster 1) by ABA. From these associations, hypotheses and predictions about the regulatory mechanisms can be proposed at work as follows. The salt-regulated clusters that show peak expression at 1 hour (Figure 10, clusters 6, 13 and 24) were associated with ABA sub-modules showing rapid activation by hormone treatment at 1 hour, whereas salt-regulated clusters showing peak expression at 3 hours (Figure 10, cluster 3 and 5) are associated with ABA sub-modules induced at 3 hours. So the first prediction is that ABA was responsible for the early induction of transcriptional responses, and that ABA biosynthesis was induced at a very early stage during salt stress response. On the other hand, salt-regulated cluster 1, which was induced late in the time course, was associated with an ABA-repressed sub-module (Figure 9E). The second prediction is that ABA might prevent the early activation of genes in cluster 1. Since the cytokinin induced module was also associated with cluster 1, it can also be predicted that cytokinin was responsible for the reactivation of the genes at late stage and that cytokinin signaling was inhibited by ABA at early stage. 69

87 To test the above predictions of what essential roles ABA signaling played in these transcriptional responses, I examined the effect of the ABA biosynthetic mutation aba2 on salt regulated expression (Gonzalez-Guzman et al., 2002). Firstly, I found that many genes that peaked at 1 hour or 3 hours after salt treatment were inhibited in their expression in aba2 mutants (Figure 10). This is consistent with our first prediction. Then it was also able to be observed clear hyper-activated expression of those salt repressed genes in cluster 1 in the aba2 mutant background between 1 and 3 hours, as predicted, although the qpcr analysis of wild-type roots was not able to reproducibly detect the late salt-mediated activation of these genes (Figure 10). Together these data showed that ABA signaling was essential for promoting expression of genes early in the salt stress time course and for preventing precocious activation of late expressed genes. To test if there was a crosstalk between ABA and cytokinin in early stage, I checked the expression of the cytokinin induced module in aba2 mutant background also, and found that the early repression was released, and this effect also happened (Figure 11). Finally, ABA levels during salt stress were quantified by the previous research assistant, Cliff Tham, in whole roots using LC-MS (Figure 12C). A dramatic increase was observed in ABA amount that peaked between 3 and 8 hours of salt treatment (Figure 12C). The initial degradation product of ABA, PA, showed a similar trend of accumulation, while DPA, which is the final catabolic product of ABA, rose thereafter and peaked by 24 hours (Figure 12C) (Finkelstein et al., 2002), indicating that ABA biosynthesis was dynamically regulated with peak levels associated with the early-phases of the salt stress response, consistent with our prediction. Furthermore, both q-rt PCR using the whole roots and our spatiotemporal map showed the quick induction of the ABA biosynthesis genes, NCED3 and ABA1 (Figure 12A&C). 70

88 Figure 10. ABA plays roles in early stage of salt stress response. Expression of genes in ABA activating sub-module and ABA repressing sub-module were detected in aba2 mutant, which is defective in ABA biosynthesis. Three time points (1, 3, and 32 hours) were used for the salt treatment. 71

89 Figure 11. Potential crosstalk between ABA and Cytokinin for the regulation of the gene expression at early stage. Genes in the cytokinin induced sub-module is reactivated in aba2 mutant at early stage, indicating that there is a crosstalk between ABA and cytokinin in early stage of salt stress response. 72

90 Figure 12. ABA biosynthesis is regulated in early stage of salt stress response. (A) The ABA biosynthetic pathway is illustrated, and efp representations show that the expression of associated genes significantly differentially expressed during salt stress. (B) Q-RT PCR detection of the ABA biosynthetic genes in salt and salt/aba conditions during early and late stages. (C) Quantifications of the ABA and ABA metabolites level during the first 24 hours of salt stress response. * indicates significance compared with standard condition. *ABA quantification was conducted by Cliff Tham. 73

91 3.3.5 ABA signaling mediated transcriptional response to salt stress showed tissue specificities Based on our network, it is known that clusters 3 and 5 showed highly cell-type specific transcriptional responses, which indicated that ABA signaling was also able to control very spatially resolved downstream pathways. Endodermis-dependent ABA signaling has also been shown regulating the lateral root growth (Duan et al., 2013). To determine whether ABA signaling functions similarly in transcriptional regulation in primary root, I utilized the same strategy involving the mis-expression of the abi1-1 coding sequence to inhibit ABA signaling in specific cell layers as previously described (Duan et al., 2013). I used the GAL4-VP16/UAS transactivation system to drive a UAS::abi1-1 transgene in different cell layers of the root (Figure 13A) and monitored the impact on ABA submodules involved in salt stress using high-throughput qrt-pcr. Of the 94 genes assayed, 30 genes showed significant genotype dependent effects by environment (Figure 13B). Of the 11 genes that were up-regulated, abi1-1 expression had the greatest impact on the saltstress response when driven by the Q2500 enhancer trap, which drives strong expression in the endodermis and weaker expression in the pericycle of the primary root (Figure 13B). Several genes showed an intriguing correlation between their expression pattern based on the spatiotemporal map and the cell layers where ABA signaling was most critical for salt regulation. For example At5g14130, which encodes a protein with predicted peroxidase activity, was specifically activated in the endodermis based on our spatiotemporal map and was strongly dependent on ABA signaling in the END/PER tissue layers (Figure 13C). I also assayed the role of tissue-specific ABA signaling in controlling the expression of genes in salt cluster 1, which I predicted to be repressed by ABA early in the time course. 74

92 Interestingly, I found that several of these genes, including RPL31B, showed significant de-repression under salt stress conditions when ABA signaling is inhibited in the END/PER layers (Figure 13D-F). Unexpectedly, it was found that ABA signaling in the LRC/EPI, COR/END and END/PER was necessary to promote the expression of these genes under standard conditions. These later results are not readily interpretable, but suggest complex roles for ABA signaling in controlling protein translation under nonstress conditions. 75

93 Figure 13. Cell layer specific ABA signaling regulates spatially localized transcriptional changes. (A) Expression of UAS:erGFP reporter from the four different GAL4-VP16 enhancer trap lines used. J0951, J0571, Q2500, and Q0990 drive expression in epidermis/lateral root cap, coretex/endodermis, endodermis/pericycle, and stele. Bar = 20 um. (B) Salt responsive gene expressions were detected in control genotype and different transactivation lines expressing abi1-1 mutant protein in select tissue layers. (C) efp diagram shows expression of a peroxidase gene (At5g14130) analyzed in (B). (D) efp diagram shows expression of SNRK2.6 (At4g33950) analyzed in (B). (E) efp diagram shows expression of COBRA (At5g60920) analyzed in (B). (F) efp diagram shows expression of RPL31B (At4G26230) analyzed in (B). 76

94 3.3.6 Dynamic involvement of GA signaling during salt stress response It was also known from our network that sub-modules of GA signaling, which is an important growth hormone, were associated with salt-regulated clusters, clusters 19 and 21, which showed early transcriptional repression and late recovery during the stress response (Figure 9F), predicting the production of GA can be decreased early by salt followed by a recovery in the late stage of salt response. The ProRGA1:GFP:RGA1 reporter can be used to track GA signaling, as GA perception leads to the degradation of the GFP:RGA protein. Indeed, I observed that GFP-RGA fluorescence intensity increased between five to eight hours after salt treatment and diminished by 48 hours (Figure 14A&B), which is well correlated with the predicted dynamics of GA signaling. The reduction in GFP-RGA fluorescence late in the salt response suggested that GA signaling may partially recover at these times and play a positive role in the reactivation of growth. I tested this hypothesis by performing live imaging of salt-treated roots supplemented with paclobutrazol (PAC), a GA-biosynthesis inhibitor (Rademacher, 2000). In contrast to the control salt treatment where roots recovered growth between 7 to 10 hours after treatment, PAC treated roots were strongly inhibited in their recovery (Figure 14C). It suggested that dynamic changes in the biosynthesis of GA are critical for determining the temporal pattern of growth regulation during salt stress. 77

95 Figure 14. GA signaling was dynamically regulated during salt stress. (A) Confocal images showed the maximum intensity projection of fluorescence from a root tip expressing the ProRGA1:GFP:RGA1 reporter. Roots were transferred to standard or salt stress medium for various lengths of time. (B) Quantification of GFP intensity at different time points after salt treatment in ProRGA1:GFP:RGA1 expressing roots (n is greater than five for each condition). (C) Effect of PAC treatment on growth under standard or salt stress conditions. PAC treatment strongly inhibited the ability of the root to recover growth rates after salt treatment. 78

96 3.4 Discussions In this study, a high-resolved spatiotemporal analysis was performed to understand a high salinity response in plants from the initial moments of perception to the long-term adjustments leading to transcriptional homeostasis. By analyzing transcriptional regulation at high temporal resolution and defining transitions in biological functions using our multidimensional microarray data set, we were able to parse out several critical regulatory pathways important for the salt response. According to our analysis, these changes include cell structure, growth, and protein biosynthesis. Using a comparative analysis methodology, it is defined that hormonal signals are predicted to have the largest role in transcriptional regulation, identified their target pathways and cell types, and determined their time of action. While the importance of ABA signaling has been clearly demonstrated for salt stress and other water stress responses (Duan et al., 2013; Xu et al., 2013), our understanding of when and where these signals appear and act functions is limited. Through the analysis of our spatiotemporal map together with the network between salt and hormones, we predicted that the biosynthesis of ABA and regulation of this pathway occurs through multiple layers of time-point and tissue-specific signaling at a very early stage under salt stress response (Figure 9C). Then the quantification of ABA level and expression patterns of the biosynthetic genes according to our spatiotemporal map suggested that the ABA is induced quickly by salt stress, earlier than 3 hours (Figure 12). And ABA signaling plays important functions in regulating the target gene expressions in the very early stage (Figure 9C and 10). ABA has the property that low levels of the hormone promote growth while high levels inhibit growth (Finkelstein et al., 2002). In the previous work, ABA signaling acts to suppress growth of lateral root (Duan et al., 2013), while in this study, it is intriguing that for primary root the phase of the salt response where ABA biosynthesis 79

97 is most important for promoting growth during the recovery and homeostasis phases and much smaller differences in ABA accumulation compared with control conditions happened (Geng et al., 2013). We have directly studied the effects of tissue-specific signaling using the mis-expression of abi1-1 to inhibit the ABA pathway in a spatiallylocalized manner. These data have highlighted the endodermis as an important site for ABA signaling in the root growth, which is the same with the previous study in lateral root growth regulation (Duan et al., 2013). In addition, ABA can crosstalk with other hormones in salt stress response regulating transcriptional homeostasis. For the cytokinin induced sub-module, ABA can repress the early activation and with the ABA decreasing, these genes were reactivated (Figure 9E and 12). This suggested that in salt stress response, ABA can regulate some target genes through repressing cytokinin pathway. The confronting interaction between ABA and cytokinin has also been reported in the study of seed germination of Arabidopsis (Wang et al., 2011). Other hormones involvements in salt stress response were also identified from the network in this study. For example, GA level changes dynamically in the salt stress response and it is correlated with the primary root growth recovery to the homeostasis (Figure 9F and 14). Together, this work shows that dynamic hormonal signaling is necessary for the transcriptional regulation and the root growth regulation. These events likely require tight coordination with the regulation of other biological processes important for long-term acclimation to a saline environment. Importantly, this work also shows that no single hormonal pathway determines the complete temporal architecture of the salt stress response. It will be critical for future studies to understand the mechanisms that regulate the rate of hormone biosynthesis and catabolism and how these pathways ultimately regulate the growth of the root through cell type specific signaling. Furthermore, while our secondary signaling network has identified little transcriptional evidence of crosstalk between the hormonal pathways, posttranscriptional crosstalk mechanisms and non- 80

98 hormonal signaling pathways, such as calcium ions and reactive oxygen species, are likely to be important mechanisms for signal integration. 81

99 Chapter 4 Results and discussions II 82

100 4.1 Abstract Complex transcriptional networks composed of cis-regulatory elements (CREs) and their corresponding transcription factors (TFs) allow people to understand how higher plants are normally developed and transcriptionally respond to environmental stimuli. Although in the past numerous putative CREs were computationally predicted, only a few were experimentally verified with their biological functions. Here, I developed an efficient pipeline for generating and validating the transcriptional network that plays an important role in the salt stress response in the Arabidopsis root. The pipeline includes: bioinformatics search and functional validation of CREs, high-throughput screening of TFs binding the CREs via yeast one hybrid and the functional validation of the TFs, as well as generation of transcriptional network. Using this pipeline, I have validated the regulatory functions of seven CREs, including ABRE (ABA response element), which is known to be involved in salt and drought stresses, and two other previously unknown elements. Here, the strategy I used is novel and efficient in studying the biological functions of CREs. In addition, the parameters for this approach were tested systematically to get an optimal condition for further use. Transcription factors were screened from a 1,958 TFs library of Arabidopsis, including several bzip and C2H2 zinc finger family TFs for ABRE, and the functions of the TFs in regulating reporter expression have been validated. 83

101 4.2 Introduction Transcriptional regulation has been shown essentially involved in many biological processes in plants, leading to the normal development as well as environmental adaptions (Pascuzzi and Benfey 2009; Kaplan-Levy et al., 2012). The regulation that drives specific gene expression patterns is fundamentally at the transcription level, which is mainly mediated by the interactions between transcription factors and cis-regulatory elements in promoters. These interactions can be represented as gene regulatory networks that allow people to study insight design principles of gene control and the mechanisms of organismal development, growth, homeostasis and environmental responses. Gene regulatory network is the most common one, involving interactions between transcription factors and their target genes. In unicellular organisms, such as bacterial, transcriptional networks have been set up, providing direct context appropriate adjustments in internal physiology (Matinez-Antonio and Collado-Vides 2003). For multicellular organisms, different transcriptional networks are also generated revealing the regulation for a specific organ s development or environmental responses (Zhong et al., 2010; Ishida et al., 2008). A stele-specific gene regulatory network was generated using high spatial resolution gene expression data set, with yeast one hybrid and yeast two hybrid, and in this network it illustrated the interaction between the mirna coding genes and their regulatory transcription factors (Brady et al., 2011). Yeast two hybrid screens used to search for the interaction between TFs and the target regulated proteins. Nowadays, yeast one hybrid screens are also widely used based on the interaction between protein and specific DNA motifs, and it is a technique starting with gene promoters. For example, transcriptional networks were generated in C.elegans using this technology (Reece-Hoyes et al., 2011). Enhancers and repressors in gene promoters are the major sources of transcriptional regulation for switching on or off gene expression under specific cues. In recent years, more and more studies have focused on this topic. 84

102 The latest studies on cancer cells suggest that super-enhancers (joint short stretches of enhancers) can catalyze high production of MYC protein that cause the cancer cells outof-control growth (Loven et al., 2013; Whyte et al., 2013). In plants, there are also a series of cis-regulatory elements identified and functionally validated based on specific genes, involved in stresses, such as ABRE, L1 box, W-box, etc. But these limited elements are further away from explaining the complicated organismal system. That means more functional regulatory elements need to be identified. The progress in the bioinformatic analysis of promoter sequences and the availability of transcriptomic data sets has led to the systematic identification of a great number of putative CREs involved in normal developmental process and environmental stress. However, in these previous studies, the biological functions of the putative regulatory sequences are still not validated. Traditionally, the downstream validation of the function of these sequences is through site-directed mutagenesis of endogenous promoters or truncated promoters. This has been done on the promoters of barley ABA-responsive HVA22 gene and shown that a region named ABRE3 is very important for the gene to respond to ABA and another coupling element, CE1, is indispensable (Shen et al., 1995). However, this strategy is not appropriate for a systemic validation for the putative CRE candidates from computational analysis because of the time and difficulty in manipulation. Specifically, (1) based on the previous study of ABRE, multiple copies of a regulatory sequence can be in a promoter, so several different versions of the promoter are needed to get a positive result; (2) the inability of performing homologous recombination in Arabidopsis requires that multiple independent transgenic lines for both the wild-type and mutated reporter must be generated and compared to determine the effect of the sequence mutations on the expression patterns; (3) the information gathered from the promoter mutant studies is likely to be highly specific to the promoter being analyzed, since the effect of the context information of a promoter. As a consequence of this, synthetic 85

103 promoters containing the regulatory sequences have begun to be used recently in studying the biological function of CREs. For example, several repeats of a specific putative regulatory element have been used as a synthetic promoter to drive a reporter gene to illustrate the biological outcomes, such as elements involved in circadian clock (Michael et al., 2008) or hormone signaling (Ulmasov et al., 1997). But here the CRE used is only repeats of the CRE elements, which is not appropriate for transcription factors screening because of the length limiting the conformational access for the transcription factor binding to the motif sequence. In this study, I did fundamental work on building up a CRE-centered transcriptional network in salt stress response, composed of functionally validated elements and their corresponding transcription factors. The pipeline for generating the transcriptional network involves the synthetic promoter strategy and high through-put yeast one hybrid screening. I have validated the biological functions of a series of putative CREs derived from the bioinformatic analysis, which lead to specific expression patterns in space and dynamic salt stress response. From the yeast one hybrid screening using an exhaustive transcription factor library (1,958 Arabidopsis transcription factors) generated by our collaborator, I obtained lists of the upstream binding proteins for these CREs. The functions of ABRE binding transcription factors involved in salt stress response were also validated. 4.3 Results Schematic description of the pipeline for setting up the transcriptional network To set up a CRE centered transcriptional network involved in salt stress response, the first step is to identify putative CRE sequences, which may have regulatory functions, using 86

104 bioinformatics tools. The most common strategy is to analyze the promoters of coexpressed or co-regulated genes from a transcriptome and get the conserved motifs. Previous studies have predicted a series of putative cis-regulatory elements for biotic/abiotic stress in plants and yeast (Maruyama et al., 2012; Harbison et al., 2004). In my study, I used our spatiotemporal data set to search for elements regulating gene expression in specific cell types or time points in the salt stress response (Figure 15). Although there are many putative CREs predicted systematically in previous studies, few were validated in their biological functions due to the reasons described above. To determine whether the pcres identified are functional, a synthetic promoter approach was developed combining with high throughput yeast one hybrid screening for their upstream binding proteins (Figure 15). Briefly, the synthetic promoter is a sequence with the CRE unit multimerized six times and fused 5 of a strong minimal promoter (minimum requirement for the binding of RNA polymerase). The CRE unit was chosen from two different genes that share the same or similar expression pattern based on the microarray data. Other than the exact CRE sequences in the center of the unit, flanking sequences from the two genes were also included, making the unit 30 base pairs in total (Figure 15). The synthetic promoters I designed are flanked by Gateway recombination sites to enable rapid cloning into plant expression vectors. The sequences of synthetic promoters were sent for gene-synthesis service. Then they were cloned to drive the expression of reporters, GUS or LUC, allowing me to assay gene expression in space or temporal dynamics (Figure 15). Once the biological functions of the CREs were validated, the two different versions of multimerized CRE unit were directly cloned into a yeast expression vector, with Golden Gate Cloning strategy, as a DNA bait for the yeast one hybrid screening against a library containing about 1,958 transcription factors (Figure 15). Then the TF-pCRE interactions would be validated in planta by over-expressing the TFs that overlap between the two independent versions of synthetic promoters, in the context 87

105 of the pcre-synthetic promoter reporter. Thus a CRE centered transcriptional network will be generated based on these functionally validated transcription factors and CREs involved in spatiotemporal salt stress. 88

106 Figure 15. Schematic chart showing the workflow for the synthetic promoter approach. (A) Bioinformatics analysis is conducted on the microarray data set, and the conserved motifs enriched in the genes showing the same expression pattern are extracted. (B) The sequences of the motifs are randomly extracted from two different contexts of gene promoters with flaking sequences, and the total length of the unit is 30bp. Then the unit sequence is repeated by 6 times. The multiple units is defined as synthetic promoter. (C) The synthetic promoter is fused with a proper minimal promoter and ligated with GUS or LUC reporter genes to check the spatial pattern of the synthetic promoter and the dynamic changes under salt stress. (D) On the other hand, the synthetic promoter is directly used in the yeast one hybrid for the transcription factors binding to it. (E) The transcriptional network can be set up using the screened TFs and the genes containing the core motifs that have a specific expression pattern. 89

107 4.3.2 Identification of the salt responsive cis-regulatory elements based on the spatiotemporal transcriptional map of Arabidopsis roots To search for the pcres for specific gene expression patterns regulated by salt, I used bioinformatics tools to analyze the conserved motifs or sequences enriched in the promoters of each salt regulated cluster of genes discussed in Chapter 3 Results. The first tool I used was the website-based Athena ( which can give me the known elements or motifs gathered from the previous studies. The AGI IDs of genes in each cluster were put in for analysis. The previous study has summarized that for Arabidopsis, regulatory elements are usually localized in less than 1 or 2 kbs on promoters due to the compact organization of the genome (Riechmann 2002). Thus I analyzed the region 1kb upstream the transcription starting sites. P-values for the enrichment significance of the obtained elements to each of the cluster were recorded and a heat map was generated using the minus log10 form of the p-values (Figure 16). The other tool I used was FIRE, which was developed in Tavazoie lab (Elemento et al. 2007) and which can give rise to both known and unknown elements. In addition, the gene lists of the 25 salt clusters were input for analysis, and it generated a heat map as in Figure 17. From the results of the two analyses, it can be seen that most of the known elements obtained are related to environmental stress, such as DRE (drought responsive element), EveningElement (involved in circadian regulation), LTRE promoter motif (low temperature responsive element), I box (light-responsive element), W-box (WRKY transcription factors binding sites), and the MYC and MYB binding motifs that are widely involved in stress responses according to the previous studies (Figure 16 and Figure 17, Abe et al., 2003). This suggested that the transcriptional network involved in salt stress was the same as other different kinds of stresses through sharing of these cis- 90

108 regulatory elements. The other important category of elements obtained is those involved in hormone signaling pathways (Figure 16 and Figure 17). For example, ABRE (ABA responsive element), ATHB6 binding motif (reported as the target of ABI3 in ABA signaling pathway), EIN3 BS in ERF1 and ARF binding site motif (involved in ethylene signaling pathway), and GADOWNAT (GA repressed elements). This, to some extent, suggested the involvement of hormone in salt stress (Figure 16 and Figure 17). In addition, some unknown elements (with high enrichment) were also obtained from FIRE analysis (Figure 17), suggesting that novel cis-regulatory elements are responsible for the spatiotemporal salt regulation. The biological functions of these unknown elements are to be validated. I also found that some elements are shared by different clusters, like telobox and ABRE (or ABRE-like) which are enriched in cluster 1&9, and cluster 6&13&24 (Figure 16 and Figure 17), and these elements can "lead" to the patterns of up-regulation in different time periods under salt stress. Those clusters sharing the same regulatory elements were very similar in their expression patterns, suggesting that the cis-regulatory elements mediated a function of inducing or repressing expression in a specific tissue or time point. There are also some elements, such as MYC binding motif, ARF BS in ERF (Figure 16) and Wbox (Figure 17), that are responsible for the repression of gene expression in salt stress response. Based on this, it can be hypothesized that the elements obtained can lead to a specific expression pattern mediating spatiotemporal salt stress response. On the other hand, it was found that a cluster of genes may be enriched in several different elements, suggesting a unique expression pattern of a cluster may be the regulatory consequences of different elements (Figure16). 91

109 Figure 16. The identification of known elements enriched with the 25 salt clusters using the method of Athena. The minus log p value was used to generate the heat map. The columns represent the 25 salt clusters. The rows are different known elements in the Athena database. 92

110 Figure 17. The identification of known elements enriched with the 25 salt clusters using the method of FIRE. The columns represent the 25 salt clusters. The rows are different known and known elements obtained from the search of random 7-mer sequences with the promoters of each salt cluster. Yellow represents over-representation, and blue represents underrepresentation. Red square represents significance. 93

111 4.3.3 Synthetic promoters harboring CREs confer the ability to drive specific expression patterns under normal or stress conditions Having these predicted putative CREs, the question is how to validate the functions, especially of a comparable amount of CREs. In this pipeline, I designed the synthetic promoter strategy described above. Then the question is whether these multimerized short sequences have the ability to drive inducible expression of reporters, potentially representing the endogenous genes. In order to answer this question, I began with a simple test using some putative CREs predicted having spatial information. To illustrate, the high resolution spatial map of Arabidopsis roots (Brady et al., 2007), with 14 different root tissues marked, was analyzed, and I found three interesting clusters, whose expression pattern were epidermis, cortex, and stele (Figure 18), respectively. FIRE was used to search for the conserved motif in the promoters of genes in these 3 clusters, and the potential tissue-specific elements, Telo-box, MYC binding motif and W-box were obtained (Figure 18). All three elements can confer expression that partially or identically overlaps with the expression pattern of the endogenous genes (Figure 19). Specifically, compared with minimal promoter only, MYC binding motif from the gene contexts can completely drive the same expression pattern although in stele the extent is a little different among different lines (Figure 19). It can be clearly seen that in the meristem and elongation zone, expression is all over the root tip including all four cell types, but it is much stronger from the layer of cortex inside. For the outer 2 layers, the expression pattern of the GUS reporter is almost the same with the pattern I obtained from the microarray analysis showing the peak expression in cortex (Figure 19). Telo-box from the 2 gene contexts can highly drive the similar expression pattern in stele in maturation zone (Figure 19), beginning from the elongation zone, and this pattern was also consistent with microarray analysis showing the peak expression in stele (Figure 19). Likely, w-box from 94

112 both gene contexts was also consistent in driving the pattern of epidermal expression, which is the same with the endogenous genes (Figure 19). Combining the above analyses suggests that this synthetic promoter approach is effective and useful for functional validation of elements, although a very short context of sequences are included. Since only the core 6-7 bp CRE sequence is the same between the two different versions of synthetic promoter, the expression pattern is probably the consequence of the core CRE. With this approach, a further rapid functional validation can be done for the above salt responsive elements. Several elements having representative expression patterns were used to test based on FIRE analysis (Figure 20). DRE (enriched in cluster 6) and L1-box (enriched in cluster 17), which are obtained by Athena, are also in the test list. Based on what is discussed above, ABA signaling is important in regulating gene expression during salt stress. Consistent with this, ABRE is strongly enriched in several clusters of our spatiotemporal map. Although ABRE has been shown to be ABA and stress responsive (Shen and Ho 1995; Hattori et al., 2002; Narusaka et al., 2003), very little is known regarding the role of this element in spatial regulation. So the first salt responsive element tested using this approach was the ABRE. The two versions of the ABRE synthetic promoter showed near identical expression patterns, with peak expression in QC and distal columella cells or lateral root cap (Figure 21). Weaker expression was also present in the vascular tissues in the maturation zone of the root (Figure 21). Moreover, this expression pattern driven by the ABRE synthetic promoters partially overlaps with the full length promoter of RD29A but not ABI1 (Figure 21). This indicates that some other elements may exist in the ABI1 promoter and the pattern of full length promoter expression is a consequence of interactions between different elements. I also treated roots of the two synthetic promoter reporters with salt or ABA, and an increase in expression was observed (Figure 22&23), as expected. This indicated that our synthetic 95

113 promoter strategy can be used in study of salt stress response because the short sequence was demonstrated to have the ability to respond to salt stress. The 35S minimal promoter (-90~+8) was used in my study as described previously (Benfey and Chua, 1990). To determine whether the minimal promoter can cause any extensive changes to the expression pattern, I also generated reporters with alternative minimal promoter for the RD29A version of ABRE multimerized unit. Interestingly, the expression pattern in the root tip was the same with the one fused with a 35S minimal promoter (Figure 24) but not the stele expression in maturation zone. It indicated that the expression pattern of root tip is authentic for ABRE. However, using RD29A minimal promoter, the ABRE sequence did not show any significant responsiveness to ABA or salt (Figure 24). This indicated that RD29A minimal promoter was more stringent, so that it can be less sensitive to the transcription factors regulation. Since the focus of my study is to determine if the putative CREs have any regulatory function, the 35S minimal promoter was chosen to enable as much regulatory activity to be uncovered as possible. The effect of flanking sequences on the expression pattern is another issue of this approach, so I introduced two other mutated versions of the ABRE synthetic promoter. Based on the previous study (Hattori et al., 2002), the second G (Table 1), which is the most important nucleotide for this element, was substituted with C or T (Table 1). Compared with the non-mutated ABRE, both mutations caused a loss expression in QC cells and stele in the maturation zone (Figure 25). Furthermore, the level was also decreased in the lateral root cap region, although I extended the GUS staining time for the mutated ABRE reporters. This indicated that short CRE sequences indeed had biological functions in leading to specific expression spatially, probably affecting the affinity of binding with specific transcription factors in specific regions. In addition, the repeat number in my synthetic promoter approach was a parameter to be finalized for the further functional screening of the putative CREs. Originally, I chose the 96

114 repeat number of 6 based on the study of DR5, which is an auxin indicator synthetic promoter. It was found that 7 repeats of DR5 sequence was a threshold for the auxin induction (Ulmasov et al., 1997). But to be more expansive, I also tested 3 and 9 repeats for ABRE synthetic promoter. And it was found that 3 repeats of ABRE drove an expression pattern similar with the mutated ABRE in that the expression in QC and stele in maturation zone was missing (Figure 25). Also, the expression level was greatly decreased in the lateral root cap (Figure 25). On the other hand, 9 repeats of ABRE unit showed very similar expression pattern with 6 repeats, but the expression level is higher (Figure 25). This indicated that the copy number of the CRE unit in a synthetic promoter can affect the binding affinity of transcription factors in some tissue types. To simplify the approach for more putative CREs functions screening, 6 repeats was chosen to use as a general situation. With these parameters finalized, I did functional screening on the other two unknown putative salt responsive CREs for their expression patterns under standard condition (Figure 26). It can be seen that the two unknown putative CREs can drive specific expression of the GUS reporter. ATATAAT was more ubiquitous in leading to the expression along nearly the whole root under standard conditions (Figure 26). However, the putative salt responsive element from my analysis, DRE and L1 box, are exceptional ones. The expression pattern was more variable for DRE, not only between the two different DRE versions of synthetic promoter but also among the different lines of one version (Figure 27). This indicated that the transcription factors for this element may ubiquitously exist in Arabidopsis root, and this DRE may be very sensitive to environment, causing the random binding of TFs in different tissues. For L1 box, it showed inconsistent expression patterns for the two different versions, with one version not functional (Figure 28). 97

115 Figure 18. The identification of the tissue-specific elements using the high resolution spatial map. (A) Heat map showed three tissue-specific clusters, with the peak expression in epidermis, cortex, and stele. Yellow represents high expression, and cyan represents lower expression. Columns are different genes in that cluster, and rows represent different spatial markers. The analysis was conducted on the high-resolution root map (Brady et al., 2007). (B) Motifs enriched in the three tissue-specific clusters of genes. Telobox, W-box, and myc binding motif are enriched in stele, epidermis, and cortex, respectively. 98

116 Figure 19. Synthetic promoters have the ability of driving specific expression patterns. (A) The expression patterns of 35S minimal promoter in meristem, elongation zone, and maturation zone under standard condition. (B) The expression patterns of two versions of Myc binding motif, showing expression in the inner cell layers under standard condition. (C) The expression patterns of two versions of Telo box, showing specific expression in the stele in maturation zone under standard condition. (D) The expression patterns of two versions of W-box, showing specific expression in epidermis in the meristem and elongation zone, and stele in maturation zone under standard condition. (E) Diagrams showing GUS level in the four cell layers quantified from eye. (F) Average expression pattern for genes used in study based on the spatiotemporal map, standard condition. T1 plants were analyzed. 99

117 Figure 20. Salt responsive elements for the further analysis. (A) The weight matrix of ABRE and the efp pattern of cluster 6, with which ABRE is enriched. (B) The weight matrix of W-box and the efp pattern of cluster 10, with which W-box is enriched. (C) The weight matrix of Telo box and the efp pattern of cluster 1, with which Telo box is enriched. (D) The weight matrix of AACCACT and the efp pattern of cluster 2, with which AACCACT is enriched. (E) The weight matrix of ATATAAT and the efp pattern of cluster 19, with which ATATAAT is enriched. 100

118 Figure 21. The expression of ABRE synthetic promoter. (A) Expression pattern of 35S minimal promoter under standard condition. (B) Expression pattern of two different versions of ABRE, with a specific expression in the lateral root cap and QC cells and stele expression in the maturation zone under standard condition. (C) Expression patterns of the two full length promoters containing the element of ABRE under standard condition. T1 plants were analyzed. 101

119 Figure 22. Synthetic promoter confers the ability of responding to environmental stresses. (A) GUS staining of 35S minimal promoter reporter plants growing on standard or plus 140mM NaCl and 50uM ABA for 3 hours. (B) and (C) GUS staining of two different versions of ABRE synthetic promoter reporter plants growing on standard or plus 140mM NaCl and 50uM ABA for 3 hours. T2 plants were analyzed. 102

120 Figure 23. Quantification of GUS reporter driven by ABRE synthetic promoters. (A) ABI1 version of ABRE synthetic promoter. (B) RD29A version of ABRE synthetic promoter. Three regions of roots were analyzed, root tip including the lateral cap and QC cells, meristem, and stele in maturation zone. The mean intensity of GUS color was measured in imagej in red channel. For each condition, 6~8 T3 homozygous plants were analyzed. 103

121 Figure 24. Experimental test of alternative minimal promoter instead of 35S minimal promoter. (A) GUS staining of the RD29A minimal promoter reporter plants and ABRE synthetic promoter reporter plants under standard condition. T1 plants were analyzed. (B) GUS staining of the RD29A minimal promoter reporter plants and ABRE synthetic promoter reporter plants growing on standard, and plus 140mM NaCl or 50uM ABA for 3 hours. T2 plants were analyzed. 104

122 Figure 25. Test of the effect of flanking sequences and repeat number for the synthetic promoter. (A) GUS staining of the reporter lines having 6 repeats of normal ABRE sequence unit. (B) GUS staining of the reporter lines having 6 repeats of mutated ABRE sequence unit, respectively. The nucleotide G which was reported important was mutated as C. (C) GUS staining of the reporter lines having 6 repeats of mutated ABRE sequence unit, respectively. The nucleotide G which was reported important was mutated as T. (D) GUS staining of the reporter lines having 3 repeats of normal ABRE sequence unit. (E) GUS staining of the reporter lines having 9 repeats of normal ABRE sequence unit. The analysis was done for the standard condition and T1 plants were analyzed. 105

123 Figure 26. Expression patterns of the 2 unknown salt responsive cis-regulatory elements. (A) GUS staining of 35S minimal promoter reporter plants growing on standard medium. (B) GUS staining of two different versions of AACCACT synthetic promoter reporter plants growing on standard medium. (C) GUS staining of two different versions of ATATAAT synthetic promoter reporter plants growing on standard medium. T1 plants were analyzed. 106

124 Figure 27. DRE synthetic promoters showed variable expressions under normal condition. (A) GUS staining of 35S minimal promoter reporter plants growing on standard medium. (B) Examples of GUS staining on different T1 lines of DRE synthetic promoter (the version of TIR1) reporter plants growing on standard medium. (C) Examples of GUS staining on different T1 lines of DRE synthetic promoter (the version of RD29A) reporter plants growing on standard medium. 107

125 Figure 28. L1 box showed different expression patterns between the two different versions of synthetic promoters. (A) GUS staining of 35S minimal promoter reporter plants growing on standard medium. (B) GUS staining on two different versions of L1 box synthetic promoter reporter plants growing on standard medium. T1 plants were analyzed. 108

126 4.3.4 Synthetic promoters containing CREs confer the ability to respond to salt stress in a dynamic manner I wanted to determine for this synthetic promoter approach whether the synthetic promoters have the ability of responding to environmental stress in a dynamic manner so that it can be used in the further study. To answer this question, I first generated Luciferase reporter lines using the ABRE, which has been studied showing ABA and salt stress response. Figure 29 shows the experimental design I used in the analysis. In the experiment, T2 plants from 2 different single insertion lines for each version of ABRE::LUC reporter plants were analyzed. After growing on standard conditions for 5 days, the plants were transferred to MS medium plus 140mM NaCl for three different periods of time. The time for salt treatment were selected based on the spatiotemporal map, that is, a peak expression time point plus two other time points in the cluster the putative CREs are enriched with. For example, 3, 24 and 48 hours of salt treatment were used for ABRE which is enriched in cluster 6 (Figure 20). Plants were also transferred to standard condition as a control. Staggered transfer was conducted so that the plants were imaged at the same time to eliminate any circadian effect. On each plate, the UBQ10::LUC plants were used as an internal control for the effect of salt on the enzymatic reaction between Luciferase and the substrate luciferin (Figure 30). For the quantification, the mean intensity of LUC luminescence of the whole root was measured with Image J for each transgenic positive plant, and was normalized to the averaged mean intensity of LUC luminescence of UBQ10::LUC on each plate. Then the average and the statistical errors were calculated. The 35S minimal promoter did not have a significant ability to drive expression of LUC reporter either under standard condition or salt conditions (Figure 31A). However, the addition of the 6 repeats of 30bp-ABRE synthetic promoters conferred expression of the 109

127 reporter not only under standard condition but also under salt stress conditions (Figure 31A). Consistent with the above GUS staining result, the responses to salt stress of the two different versions of ABRE synthetic promoters is different, and the sequences from ABI1 gene context is more responsive (Figure 31A), suggesting extra information can be included that is important for salt response in this synthetic promoter, which needs to be further elucidated. From the luminesce images, it seemed that intensity increased with the time of salt treatment, but the intensity of UBQ10 was also increased, suggesting that salt can affect the enzymatic reaction in different stages. From the quantified data (Figure 31 B&C), the peak expression of ABRE synthetic promoter occurred at 3 hours after salt treatment, although in the RD29A version, one of the lines showed a deviation. This dynamic response is consistent with the salt responsive expression of the full length promoter based on our spatiotemporal map (Figure 31 D&E). So this result indicates that the synthetic promoters have the properties of full length promoters, driving a specific salt responsive expression pattern in a dynamic manner. In addition, the other salt responsive elements were also tested for dynamic responsiveness. First, not all the CREs synthetic promoters showed the same response to salt between the two different versions, and this was also indicated from the analysis of ABRE. For the unknown element AACCACT, the two versions of synthetic promoters showed very similar response profiles under salt stress, which was well represented with the full length promoters (Figure 33). For the other unknown element ATATAAT, no significant response was detected (Figure 33). Secondly, the different versions of the same synthetic promoter can have different dynamics of response to salt stress and they are not able to represent the full length promoters pattern. For example, for telobox, one version showed a dynamic response with the peak expression 3 hours, while the other version showed a down-regulation in the whole process of salt stress response (Figure 32). Third, for some specific version of synthetic promoter, little variation in the dynamic 110

128 pattern between independent lines was seen. Wbox synthetic promoter showed a very consistent pattern between different lines (Figure 32) that is the same as the full length promoter, while for DRE and Telobox, one of the versions showed a very consistent pattern, while the other version showed variations in different lines (Figure 32). Together, although most of the synthetic promoters have the ability to dynamically respond to salt stress, they have different activities compared to the full length promoter. The reason could be that the process of responding to salt is very complicated and needs combinatorial elements to mediate a specific pattern. Also, although the core motif sequences are enough to determine the similar profiles of synthetic promoters, the analysis using LUC lines here suggested variations of synthetic promoters in temporal profile responding to salt stress. Another reason is that sometimes use of a reporter system will change the temporal dynamics by observation compared to the direct measurement of RNA concentration by microarray. 111

129 Figure 29. Experimental design for the dynamic response of synthetic promoters under salt stress. Seeds are germinated on MS condition and grown for 5 days. Then the seedlings were transferred to standard medium and plus 140mM for different time of treatment. The times were used as 3, 24, and 48 hours based on the spatiotemporal map. 1 hour before imaging, the seedlings were treated with 1mM luciferin. 112

130 Figure 30. efp showing the spatiotemporal expression pattern of UBQ10 under salt stress reponse. UBQ10 does not show any significant changes at transcriptional level under salt stress, so the UBQ10::LUC reporter plants can be used as internal control for the effect of salt and different periods of time on enzymatic reactions of luciferase and the substrate luciferin. 113

131 Figure 31. ABRE synthetic promoters respond to salt stress dynamically. (A) Luminescence images of two versions of ABRE::LUC under salt treatment for 3, 24, and 48 hours. On each plate, UBQ10::LUC plants were used as an internal control, and based on the spatiotemporal map, UBQ10 is not responsive to salt stress. All the luminescence images were adjusted at the same scale. C, control plants. (B) Quantification of LUC intensity for the ABRE (RD29A) version synthetic promoter. 10 seedlings for each line were analyzed, and the values used here is normalized mean intensity to the averaged LUC intensity of the UBQ10::LUC plants on each plate. (C) Quantification of LUC intensity for the ABRE (ABI1) version synthetic promoter. The data was analyzed the same way as in (B). (D) and (E) The efp showing the spatiotemporal expression patterns of the genes RD29A and ABI1 under salt stress. 114

132 Figure 32. Dynamic analysis of salt stress response of the known elements. (A) Quantification of LUC intensity for the DRE synthetic promoters. Two independent lines for both versions were tested. 10 seedlings for each line were analyzed, and the values used here is normalized mean intensity to the averaged mean LUC intensity of the UBQ10::LUC plants on each plate. (B) Spatiotemporal expression patterns of full length promoters of RD29A containing DRE. (C) Spatiotemporal expression patterns of full length promoters of TIR1 containing DRE. (D) Quantification of LUC intensity for the Telobox synthetic promoters. Two independent lines for both versions were tested. Data was analyzed in the same way in (A). (E) Spatiotemporal expression patterns of full length promoters of AT5G61030 containing Telobox. (F) Spatiotemporal expression patterns of full length promoters of AT4G15770 containing Telobox. (G) Quantification of LUC intensity for the Wbox synthetic promoters. Three independent lines for one version were tested. Data was analyzed in the same way in (A). (H) Spatiotemporal expression patterns of full length promoters of AT1G22500 containing Wbox. 115

133 Figure 33. Dynamic analysis of salt stress response of the unknown elements. (A) Quantification of LUC intensity for the AACCACT synthetic promoters. Two independent lines for both versions were tested. 10 seedlings for each line were analyzed, and the values used here is normalized mean intensity to the averaged mean LUC intensity of the UBQ10::LUC plants on each plate. (B) Spatiotemporal expression patterns of full length promoters of AT1G13600 containing AACCACT. (C) Spatiotemporal expression patterns of full length promoters of AT5G23920 containing AACCACT. (D) Quantification of LUC intensity for the ATATAAT synthetic promoters. Two independent lines for both versions were tested. Data were analyzed in the same way in (A). (E) Spatiotemporal expression patterns of full length promoters of AT1G14280 containing ATATAAT. (F) Spatiotemporal expression patterns of full length promoters of AT5G53250 containing ATATAAT. 116

134 4.3.5 Synthetic promoter strategy for screening using the TF library For functionally validated cis-regulatory elements, synthetic promoters can also be used to directly screen the upstream binding transcription factors. To understand whether different parameters of the synthetic promoters can have significant effect on the robust binding of transcription factors, I tested different versions of ABRE synthetic promoters. Firstly, I hypothesized that the binding affinity of transcription factors to the motif could be affected by the DNA conformations of the repeats in synthetic promoters. Sometimes the bending of repeat DNA is with 10-base pair periodicity, so 30bp and 25bp for the repeated unit in synthetic promoters were used to test the effect of conformational folding on the binding affinity of TFs. From Figure 34A&B, it shows that 30 bp of the unit could bind more TFs. And compared with the 25bp version of the repeated unit, 30 bp usually shows higher affinity for TF binding (Figure 36 A&B). This suggested that 30 bp is a better parameter for the yeast one hybrid screening of the synthetic promoters because of its higher activity in binding TFs. In addition, I also hypothesized that the repeat number of the unit could also affect the binding specificity and affinity of TF. To test this, two repeat number, 3 and 6 were tested. The number of TFs screened out is not significant between 3 repeats and 6 repeats, although 6 repeats can give slightly more information (Figure 34 C&D). For the affinity of TF binding, it showed a little difference when combining with different length of the unit. Combining with 30 bp of the unit, 6 repeats showed higher affinity for some TFs, while combing with 25 bp of the unit, 3 repeats showed higher affinity for some TFs (Figure 36 C&D). It suggested that compared with the parameter of unit length repeat number is not a decisive parameter for the TF binding of the synthetic promoters. So, the combination of 6 repeats and 30bp fragment is a good choice for the further screening parameters of synthetic promoters, and this is consistent with the functional study as well. 117

135 As shown above, the response of the two different ABRE synthetic promoters to salt stress was different (Figure 23), that is, the RD29A version showed an obvious induction to salt only at the region of root tip, while the ABI1 version showed a much broader induction to salt. I hypothesized that the difference in flanking sequences between the two versions of synthetic promoter caused the different responses. In order to test this hypothesis, I firstly did PCA analysis using the binding affinities of the whole list of transcription factors (~2,000 TFs in Arabidopsis) from yeast one hybrid on these different test versions of ABRE synthetic promoters from both ABI1 and RD29A context background. With the first principle component of %, the two kinds of flanking sequences from different context background were significantly separated (Figure 35), suggesting the flanking sequences of the core regulatory motifs indeed have potential effects on transcription factor binding. Then I looked back at the exact flanking sequences, and for the ABI1 version synthetic promoter, there is a DRE core motif sharing some nucleotides with ABRE (Figure 35). The previous study has suggested that the DRE could mediate stress response in an ABA independent pathway, and from my result this could be region dependent. This can explain the different salt response ability of the two different ABRE synthetic promoters with different flanking sequences described above. Second, the fold induction of the overlapped transcription factors was compared between different test versions of ABRE synthetic promoters. Using the combination of 30bp and 6 repeats, I got the overlapped list of transcription factors between ABRE synthetic promoters of ABI1 version and RD29A version, shown in Table 4. Among them, 3 are from bzip family and two are from C2H2 zinc finger family. The BZIP transcription factor ABF3 at the top of the list with a high binding affinity has been reported to bind to ABRE and is involved in drought and salt tolerance. This suggests that the list of transcription factors with our filtering strategy is reasonable. GBF2 was first characterized by screening using G-box, which is similar to ABRE. It was 118

136 reported to be involved in environmental pathways also, such as light induced signaling. BZIP3 is a transcription factor that is less studied, and this is the first time that it has been shown be involved in ABA signaling pathway through binding ABRE. The further exploration of BZIP3 in the interactome database suggests that it can interact with several calmodulin ( suggesting it can be involved in salt stress and ABA signaling through calcium signaling. Otherwise, the two C2H2 zinc finger transcription factors, STZ and AZF3, were reported binding the sequence of A (G/C) T repeats, which is also similar to ABRE. These two transcription factors were also involved in high salinity tolerance, suggesting they can play the function through regulating the ABRE contained target genes. To validate whether these transcription factors obtained from yeast one hybrid screening also have the functions in plants through ABRE, I over-expressed several transcription factors in the ABRE synthetic promoter reporter background (Figure 37). Under standard condition, the ABF3 did not have an obvious function on the synthetic promoter, while AZF3 was kind of repressing the expression pattern (Figure 37). Under salt and ABA conditions, the expression of the reporter is strengthened. It indicated that the transcription factors could function as activators under stress conditions. To test whether the transcription factors are functional under salt stress, I checked the stabilization of the over-expressed proteins of the transcription factors in early and late stages of salt stress. Both of the bzip transcription factors showed a salt induced stabilization, but in different stages. ABF3 began to be stabilized by salt early, which is around 4 hours after salt treatment, and the stabilization increased with the salt treatment time prolonged to 24 hours (Figure 38A). BZIP3 was stabilized by salt mainly at early stage of 4 hours after salt treatment, and at 24 hours of salt treatment the nuclear stabilization is decreased (Figure 38B). In addition, the localization of the induced 119

137 proteins for both transcription factors is lateral root cap (Figure 38), which is consistent with the expression pattern of ABRE synthetic promoters (Figure 21). This indicated that the two transcription factors play functions in salt stress through ABRE in lateral root cap. On the other hand, the C2H2 transcription factor, AZF3, was stabilized under standard conditions in the lateral root cap. In elongation and maturation zones, the protein was significantly stabilized in the epidermis and cortex cell layers in 4 hours after salt treatment and dynamically decreased in the late stage of salt stress response (Figure 39). 120

138 Figure 34. Experimental test of different versions of ABRE synthetic promoters for TF screening using Y1H. (A) Overlap of TFs between 25bp and 30bp of the fragment with the same repeat number of 6. (B) Overlap of TFs between 25bp and 30bp of the fragment with the same repeat number of 3. (C) Overlap of TFs between 3 repeats and 6 repeats of the 30bp fragment. (D) Overlap of TFs between 3 repeats and 6 repeats of the 25bp fragment. ABI1 version was used as an example. 121

139 A B RD29A cagacgcttcatacgtgtccctttatctct DRE/CRT Core motif ABI1 ttttcttcgtctacgtgtcgaccatccacc RCCGAC (R=G/A) Figure 35. Effect of flanking sequence. (A) Principal component analysis showing the flanking effect of ABREs according to the binding affinity of the transcription factors. (B) The flanking sequences of the two versions of ABRE. The nucleotides in red are potential DRE sequences. R, synthetic promoters with RD29A context information; A, synthetic promoters with ABI1 context information. 30 and 25, 30bp and 25bp of the repeated unit. 3 and 6, 3 and 6 repeats of the unit. The fold induction values of all transcription factors analyzed were used. The first principle component is %, differentiating the ABRE synthetic promoters into different context information of the flanking sequences. And the second principle component is only %, suggesting different test versions of ABRE synthetic promoter having the same flanking information are not significantly different. This was generated using the whole list of ~1958 TFs with their fold induction values. 122

140 Figure 36. TF binding affinity comparison between different test versions of ABRE. (A) Comparison between 25bp and 30bp under the same repeat number of 6. (B) Comparison between 25bp and 30bp under the same repeat number of 3. (C) Comparison between 3 repeats and 6 repeats under the same length of fragment 30bp. (D) Comparison between 3 repeats and 6 repeats under the same length of fragment 25bp. ABI1 version was used as an example. The overlapped TFs between each pair showing a fold induction above 2 (normalized to the control TF) in yeast one hybrid screening were used for analysis. The fold induction represents TF binding affinity. X axis, the overlapped transcription factors analyzed; y axis, fold induction values. 123

141 Table 4. Transcription factors showing overlap between the two versions of synthetic promoters from Y1H screening. TFs DNA bait Gene name TF family AT4G34000 ABRE ABF3 bzip AT4G01120 ABRE GBF2 bzip AT5G15830 ABRE ATBZIP3 bzip AT1G27730 ABRE STZ C2H2 AT5G43170 ABRE AZF3 C2H2 AT2G17950 ABRE WUS HB AT3G10000 ABRE EDA31 Trihelix AT2G01370 ABRE GeBP AT2G40620 Telobox T2P4.3 bzip AT3G11260 MYC binding WOX5 HB *Fold induction>2 124

142 Figure 37. In vivo validation of the interaction between ABRE synthetic promoter and the transcription factors obtained from yeast one hybrid. ABRE::GUS reporter plants, with ABF3 or AZF3, and no TF over-expression were grown under standard conditions, and then transferred onto standard medium plus 140mM NaCl for 5 and 24 hours or standard medium plus 50uM ABA for 3 hours. Then GUS staining was conducted on the roots of the plants. DIC images were taken by compound microscope. Homozygous plants for both transgenes were analyzed. 125

143 Figure 38. Protein stabilization of BZIP family transcription factors under salt stress. (A) ABF3 was stabilized continuously with time of salt treatment increased in the lateral root cap. Plants representing here are in ABRE synthetic promoter RD29A version. The other version showed the same result. (B) BZIP3 was stabilized dynamically under salt stress response in the lateral root cap. Plants shown here are in ABRE synthetic promoter RD29A version. The other version showed the same result. 126

144 Figure 39. Protein stabilization of C2H2 family transcription factor, AZF3, under salt stress. AZF3 s stabilization induced by salt stress happened in epidermis and cortex in elongation and maturation zones. The time for the stabilization is 4 hours after salt treatment. Plants shown here are in ABRE synthetic promoter RD29A version. The other version showed the same result. 127

145 4.4 Discussion More and more studies focus on the functions of noncoding DNA sequences in regulating gene expressions. Cis-regulatory element sequences located in gene promoters are an important part of the noncoding regulatory sequence. In this study, a synthetic pipeline was first set up to understand the biological functions of short CREs sequences and the mechanism of transcription regulation in the spatiotemporal control of the salt stress response. The activity of the ABRE was first characterized which can be an indicator of the function of ABA signaling in root development. On the other hand, our analysis suggests that under environmental stress, a specific single regulatory element is not sufficient to mediate the response. The combinatorial sequences flanking the core motif are needed Synthetic promoters drive tissue-specific and salt responsive patterns In plant development and environmental response, regulation at the transcriptional level is indispensable. Cis-regulatory elements located in gene promoters can be bound by specific transcription factors, switching on or off the gene expression. Synthetic promoter strategy has been used to reveal the hormonal response, and the reporter lines of several synthetic promoters have been widely used as hormone indicators, such as DR5::GUS (Ulmasov et al., 1997). The characterization and validation of these CREs are all based on a specific gene. For example, DR5 was characterized from the promoter of soybean gene GH3 (Liu et al., 1994). The ABRE s responsive property is based on a series of truncations of the RD29B promoter (Shen et al., 1995). In this study, I identified and characterized the expression patterns of several pcres using a novel strategy of synthetic promoters. The same expression pattern led by different versions of synthetic promoters that were from different gene contexts demonstrated the biological functions of the CREs 128

146 in development of the root. Of the 8 putative CREs that I tested, 6 showed the same expression pattern between the two different versions that have different flanking information. And 3 elements showed good correlation with the predicted pattern according to microarray analysis. To some extent, the pattern driven by the short synthetic promoters can represent the full length promoter pattern (Figure 21). Using this strategy, more tissue-specific synthetic promoters can be validated from the bioinformatics analysis. And this can provide a powerful tool for engineering promoters and the spatial study of interesting proteins in root development ABRE s expression pattern indicates the location of ABA signaling in root development and environmental response ABA signaling is widely involved in root development and stress response, and ABA is usually called stress hormone due to its biological functions. ABRE is an important cisregulatory element, mediating ABA signaling in the regulation of target genes in salt stress response (Shen et al., 1995). However, the spatial information for this element has never been characterized. In this study, I revealed using ABRE synthetic promoters that ABRE led a specific expression in QC cells and lateral root cap/columella in meristem and stele in maturation zone (Figure 21) under standard condition. Zhang et al. (2010) demonstrated that ABA plays an important role in maintaining the quiescence of the quiescent center in Arabidopsis primary root development. However, the direct evidence that ABA signaling is in the QC was lacking. Our work to some extent provided an evidence of the spatial information of ABA signaling, and it suggested that ABA can maintain the root stem cell niche through regulating ABRE containing genes. Additionally, lateral root cap and stele are also locations that ABA signaling happens as a 129

147 default state under normal developmental condition. But the function of ABA in these cell types is still not clear Combinatorial properties of regulatory elements necessary for environmental stress response Although all cell types contain the same genetic information, different genes are activated in different cell types. Our synthetic promoter pipeline can well explain this issue. However, according to previous bioinformatics predictions (Figure 16 &17, Zou et al., 2011), some elements tend to co-occur in a specific region of promoters, suggesting that transcriptional regulation should be combinatorial for some biological processes. Work on cancer cells has revealed that enhancers are more powerful when they are joined together in regulating cancer cell growth (Whyte et al., 2013; Loven et al., 2013). In Drosophila mesoderm development, the precise patterns of gene expression are decided by the temporal and combinatorial binding of the TFs and cis-regulatory elements (Zinzen et al., 2009), and combinatorial regulation are highly conserved across species (He et al., 2011). In the study of environmental response in Arabidopsis, the co-existence of ABRE and DRE in the RD29A promoter confers salt and drought response, and our analysis (Figure 16) suggests the same conclusion. In my study, the two ABRE synthetic promoters showed different responses to ABA and salt treatment. The flanking sequences were analyzed and it was found that in the flanking sequence of ABI1 version, there is a potential DRE core motif (Table 1). I hypothesize that this potential DRE motif confers the ABRE synthetic promoter (ABI1 version) greater responsiveness to environmental stress. So compared with the tissue specificity of synthetic promoter, the response to environmental stress need more information beyond the core motif. 130

148 Chapter 5 Conclusions 131

149 This thesis systematically studied the spatiotemporal control of salt stress induced transcriptional responses in Arabidopsis roots. The conclusions are summarized as follows: 1) By analyzing transcriptional regulation at high temporal resolution and defining transitions in biological functions using our multidimensional microarray data set, we were able to parse out several critical regulatory pathways important for the salt response. The changes include cell structure, growth, and protein biosynthesis. 2) Using a comparative analysis methodology, it is defined that hormonal signals are predicted to have the largest role in transcriptional regulation, identified their target pathways and cell types, and determined their time of action. 3) A synthetic pipeline was set up to study the functions of CREs in regulating tissue specificity and salt responsive gene expressions. The synthetic promoters were tested having the properties of full length promoters and had the ability of binding functional transcription factors. And this pipeline is useful for generation of CRE centered transcriptional network in salt stress response. 132

150 References Abe, H., Urao, T., Ito, T., Seki, M., Shinozaki, K., & Yamaguchi-Shinozaki, K. (2003). Arabidopsis AtMYC2 (bhlh) and AtMYB2 (MYB) function as transcriptional activators in abscisic acid signaling. The Plant Cell Online, 15(1), Abogadallah, G. M. (2010). Insights into the significance of antioxidative defense under salt stress. Plant signaling & behavior, 5(4), Abràmoff, M. D., Magalhães, P. J., & Ram, S. J. (2004). Image processing with ImageJ. Biophotonics international, 11(7), Achard, P., Cheng, H., De Grauwe, L., Decat, J., Schoutteten, H., Moritz, T.,... & Harberd, N. P. (2006). Integration of plant responses to environmentally activated phytohormonal signals. Science, 311(5757), Ali, B., Hasan, S. A., Hayat, S., Hayat, Q., Yadav, S., Fariduddin, Q., & Ahmad, A. (2008). A role for brassinosteroids in the amelioration of aluminium stress through antioxidant system in mung bean (Vigna radiate L. Wilczek).Environmental and Experimental Botany, 62(2), Allakhverdiev, S. I., & Murata, N. (2008). Salt stress inhibits photosystems II and I in cyanobacteria. Photosynthesis research, 98(1-3), Alon, U. (2007). Network motifs: theory and experimental approaches. Nature Reviews Genetics, 8(6), Amtmann, A., & Sanders, D. (1999). Mechanisms of Na+ uptake by plant cells. Academic Press. 133

151 Apse, M. P., Aharon, G. S., Snedden, W. A., & Blumwald, E. (1999). Salt tolerance conferred by overexpression of a vacuolar Na+/H+ antiport in Arabidopsis. Science, 285(5431), Ashburner, M., Ball, C. A., Blake, J. A., Botstein, D., Butler, H., Cherry, J. M.,... & Sherlock, G. (2000). Gene Ontology: tool for the unification of biology. Nature genetics, 25(1), Atanassova, L., Stojanov, I., Pissarska, M., & Valkova, C. (1997). Salt stress-induced changes of cytokinins in maize and pea plants RNA. Bulg J Plant Physiol, 23, Bai, M. Y., Shang, J. X., Oh, E., Fan, M., Bai, Y., Zentella, R.,... & Wang, Z. Y. (2012). Brassinosteroid, gibberellin and phytochrome impinge on a common transcription module in Arabidopsis. Nature cell biology. 14 (8): Banu, M. N. A., Hoque, M. A., Watanabe-Sugimoto, M., Matsuoka, K., Nakamura, Y., Shimoishi, Y., & Murata, Y. (2009). Proline and glycinebetaine induce antioxidant defense gene expression and suppress cell death in cultured tobacco cells under salt stress. Journal of Plant Physiology, 166(2), Barabási, A. L., & Oltvai, Z. N. (2004). Network biology: understanding the cell's functional organization. Nature Reviews Genetics, 5(2), Barabási, A. L., & Oltvai, Z. N. (2004). Network biology: understanding the cell's functional organization. Nature Reviews Genetics, 5(2), Bartels, D., & Sunkar, R. (2005). Drought and salt tolerance in plants. Critical Reviews in Plant Sciences, 24(1), Benfey, P. N., & Chua, N. H. (1990). The cauliflower mosaic virus 35S promoter: combinatorial regulation of transcription in plants. Science, 250(4983),

152 Benfey, P. N., & Scheres, B. J. G. (2000). Root development. Current Biology,10, Benfey, P. N., & Weigel, D. (2001). Transcriptional networks controlling plant development. Plant physiology, 125(1), Bohnert, H. J., Nelson, D. E., & Jensen, R. G. (1995). Adaptations to environmental stresses. The plant cell, 7(7), Boudsocq, M., Willmann, M. R., McCormack, M., Lee, H., Shan, L., He, P.,... & Sheen, J. (2010). Differential innate immune signalling via Ca2+ sensor protein kinases. Nature, 464(7287), Brady, S. M., Orlando, D. A., Lee, J. Y., Wang, J. Y., Koch, J., Dinneny, J. R.,... & Benfey, P. N. (2007). A high-resolution root spatiotemporal map reveals dominant expression patterns. Science, 318(5851), Brady, S. M., Zhang, L., Megraw, M., Martinez, N. J., Jiang, E., Yi, C. S.,... & Benfey, P. N. (2011). A stele-enriched gene regulatory network in the Arabidopsis root. Molecular systems biology, 7(1): 459. Brown, C. E., Howe, L., Sousa, K., Alley, S. C., Carrozza, M. J., Tan, S., & Workman, J. L. (2001). Recruitment of HAT complexes by direct activator interactions with the ATM-related Tra1 subunit. Science, 292(5525), Bruex, A., Kainkaryam, R. M., Wieckowski, Y., Kang, Y. H., Bernhardt, C., Xia, Y.,... & Schiefelbein, J. (2012). A gene regulatory network for root epidermis cell differentiation in Arabidopsis. PLoS genetics, 8(1), e Burssens, S., Himanen, K., Van de Cotte, B., Beeckman, T., Van Montagu, M., Inzé, D., & Verbruggen, N. (2000). Expression of cell cycle regulatory genes and 135

153 morphological alterations in response to salt stress in Arabidopsis thaliana. Planta, 211(5), Cao, S., Xu, Q., Cao, Y., Qian, K., An, K., Zhu, Y.,... & Kuai, B. (2005). Loss offunction mutations in DET2 gene lead to an enhanced resistance to oxidative stress in Arabidopsis. Physiologia Plantarum, 123(1), CAO, W. H., Liu, J., ZHOU, Q. Y., CAO, Y. R., ZHENG, S. F., DU, B. X.,... & CHEN, S. Y. (2006). Expression of tobacco ethylene receptor NTHK1 alters plant responses to salt stress. Plant, cell & environment, 29(7), Cao, Y. R., Chen, S. Y., & Zhang, J. S. (2008). Ethylene signaling regulates salt stress response: An overview. Plant signaling & behavior, 3(10), Carter, C. T., Grieve, C. M., & Poss, J. A. (2005). Salinity effects on emergence, survival, and ion accumulation of Limonium perezii. Journal of Plant Nutrition, 28(7), Chaves, M. M., Flexas, J., & Pinheiro, C. (2009). Photosynthesis under drought and salt stress: regulation mechanisms from whole plant to cell. Annals of Botany, 103(4), Chen, L. T., Luo, M., Wang, Y. Y., & Wu, K. (2010). Involvement of Arabidopsis histone deacetylase HDA6 in ABA and salt stress response. Journal of experimental botany, 61(12), Cheng, Y., Dai, X., & Zhao, Y. (2007). Auxin synthesized by the YUCCA flavin monooxygenases is essential for embryogenesis and leaf formation in Arabidopsis. The Plant Cell Online, 19(8),

154 Chinnusamy, V., Schumaker, K., & Zhu, J. K. (2004). Molecular genetic perspectives on cross talk and specificity in abiotic stress signalling in plants.journal of Experimental Botany, 55(395), Chrispeels, M. J., & Agre, P. (1994). Aquaporins: water channel proteins of plant and animal cells. Trends in biochemical sciences, 19(10), Clough, S. J., & Bent, A. F. (1998). Floral dip: a simplified method foragrobacteriummediated transformation ofarabidopsis thaliana. The Plant Journal, 16(6), Cosma, M. P., Panizza, S., & Nasmyth, K. (2001). Cdk1 triggers association of RNA polymerase to cell cycle promoters only after recruitment of the mediator by SBF. Molecular cell, 7(6), Dassanayake, M., Oh, D. H., Haas, J. S., Hernandez, A., Hong, H., Ali, S.,... & Cheeseman, J. M. (2011). The genome of the extremophile crucifer: Thellungiella parvula. Nature genetics, 43(9), Davenport, R. J., MUÑOZ MAYOR, A. L. I. C. I. A., Jha, D., Essah, P. A., Rus, A. N. A., & Tester, M. (2007). The Na+ transporter AtHKT1; 1 controls retrieval of Na+ from the xylem in Arabidopsis. Plant, cell & environment, 30(4), Dearden, P., & Akam, M. (1999). Developmental evolution: Axial patterning in insects. Current biology, 9(16), R591-R594. Deplancke, B., Dupuy, D., Vidal, M., & Walhout, A. J. (2004). A gateway-compatible yeast one-hybrid system. Genome research, 14(10b), Dinneny, J. R., Long, T. A., Wang, J. Y., Jung, J. W., Mace, D., Pointer, S.,... & Benfey, P. N. (2008). Cell identity mediates the response of Arabidopsis roots to abiotic stress. Science, 320(5878),

155 Drew, M. C., & Lynch, J. (1980). Soil anaerobiosis, microorganisms, and root function. Annual Review of Phytopathology, 18(1), Duan, L., Dietrich, D., Ng, C. H., Chan, P. M. Y., Bhalerao, R., Bennett, M. J., & Dinneny, J. R. (2013). Endodermal ABA signaling promotes lateral root quiescence during salt stress in Arabidopsis seedlings. The Plant Cell Online, 25(1), Elemento, O., Slonim, N., & Tavazoie, S. (2007). A universal framework for regulatory element discovery across all genomes and data types. Molecular cell, 28(2), Ephrussi, A., & Johnston, D. S. (2004). Seeing is believing: the bicoid morphogen gradient matures. Cell, 116(2), Farooq, M., Wahid, A., Lee, D. J., Ito, O., & Siddique, K. H. (2009). Advances in drought resistance of rice. Critical Reviews in Plant Sciences, 28(4), Finkelstein, R. R., Gampala, S. S., & Rock, C. D. (2002). Abscisic acid signaling in seeds and seedlings. The Plant Cell Online, 14(suppl 1), S15-S45. Flowers, T. J., & Colmer, T. D. (2008). Salinity tolerance in halophytes*. New Phytologist, 179(4), Flowers, T. J., Hajibagheri, M. A., & Clipson, N. J. W. (1986). Halophytes.Quarterly Review of Biology, Flowers, T. J., Troke, P. F., & Yeo, A. R. (1977). The mechanism of salt tolerance in halophytes. Annual Review of Plant Physiology, 28(1), Flowers, T., & Yeo, A. R. (1986). Ion relations of plants under drought and salinity. Functional Plant Biology, 13(1), Fujioka, S., & Yokota, T. (2003). Biosynthesis and metabolism of brassinosteroids. Annual Review of Plant Biology, 54(1),

156 Gallego-Bartolomé, J., Minguet, E. G., Grau-Enguix, F., Abbas, M., Locascio, A., Thomas, S. G.,... & Blázquez, M. A. (2012). Molecular mechanism for the interaction between gibberellin and brassinosteroid signaling pathways in Arabidopsis. Proceedings of the National Academy of Sciences, 109(33), Galvan-Ampudia, C. S., & Testerink, C. (2011). Salt stress signals shape the plant root. Current opinion in plant biology, 14(3), Gendron, J. M., Pruneda-Paz, J. L., Doherty, C. J., Gross, A. M., Kang, S. E., & Kay, S. A. (2012). Arabidopsis circadian clock protein, TOC1, is a DNA-binding transcription factor. Proceedings of the National Academy of Sciences, 109(8), Geng, Y., Wu, R., Wee, C. W., Xie, F., Wei, X., Chan, P. M. Y.,... & Dinneny, J. R. (2013). A Spatio-Temporal Understanding of Growth Regulation during the Salt Stress Response in Arabidopsis. The Plant Cell Online, doi: Ghanem, M. E., Albacete, A., Martínez-Andújar, C., Acosta, M., Romero-Aranda, R., Dodd, I. C.,... & Pérez-Alfocea, F. (2008). Hormonal changes during salinityinduced leaf senescence in tomato (Solanum lycopersicum L.). Journal of experimental botany, 59(11), Gifford, M. L., Dean, A., Gutierrez, R. A., Coruzzi, G. M., & Birnbaum, K. D. (2008). Cell-specific nitrogen responses mediate developmental plasticity. Proceedings of the National Academy of Sciences, 105(2), González-Guzmán, M., Apostolova, N., Bellés, J. M., Barrero, J. M., Piqueras, P., Ponce, M. R.,... & Rodríguez, P. L. (2002). The short-chain alcohol dehydrogenase ABA2 catalyzes the conversion of xanthoxin to abscisic aldehyde. The Plant Cell Online, 14(8),

157 Guzy, R. D., Hoyos, B., Robin, E., Chen, H., Liu, L., Mansfield, K. D.,... & Schumacker, P. T. (2005). Mitochondrial complex III is required for hypoxia-induced ROS production and cellular oxygen sensing. Cell metabolism, 1(6), Halperin, S. J., Gilroy, S., & Lynch, J. P. (2003). Sodium chloride reduces growth and cytosolic calcium, but does not affect cytosolic ph, in root hairs of Arabidopsis thaliana L. Journal of experimental botany, 54(385), Harbison, C. T., Gordon, D. B., Lee, T. I., Rinaldi, N. J., Macisaac, K. D., Danford, T. W.,... & Young, R. A. (2004). Transcriptional regulatory code of a eukaryotic genome. Nature, 431(7004), Hasegawa, P. M., Bressan, R. A., Zhu, J. K., & Bohnert, H. J. (2000). Plant cellular and molecular responses to high salinity. Annual review of plant biology, 51(1), Haseloff, J. (1998). GFP variants for multispectral imaging of living cells. Methods in cell biology, 58, Hathaway, R. J., Bezdek, J. C., & Pedrycz, W. (1996). A parametric model for fusing heterogeneous fuzzy data. Fuzzy Systems, IEEE Transactions on, 4(3), Hattori, T., Totsuka, M., Hobo, T., Kagaya, Y., & Yamamoto-Toyoda, A. (2002). Experimentally determined sequence requirement of ACGT-containing abscisic acid response element. Plant and Cell Physiology, 43(1), Hayashi, H., Mustardy, L., Deshnium, P., Ida, M., & Murata, N. (1997). Transformation of Arabidopsis thaliana with the coda gene for choline oxidase; accumulation of glycinebetaine and enhanced tolerance to salt and cold stress. The Plant Journal, 12(1),

158 He, Q., Bardet, A. F., Patton, B., Purvis, J., Johnston, J., Paulson, A.,... & Zeitlinger, J. (2011). High conservation of transcription factor binding and evidence for combinatorial regulation across six Drosophila species. Nature genetics, 43(5), Helariutta, Y., Fukaki, H., Wysocka-Diller, J., Nakajima, K., Jung, J., Sena, G.,... & Benfey, P. N. (2000). The SHORT-ROOT Gene Controls Radial Patterning of the Arabidopsis Root through Radial Signaling. Cell, 101(5), Hindt, M. N., & Guerinot, M. L. (2012). Getting a sense for signals: regulation of the plant iron deficiency response. Biochimica et Biophysica Acta (BBA)-Molecular Cell Research, 1823(9), Holmström, K. O., Somersalo, S., Mandal, A., Palva, T. E., & Welin, B. (2000). Improved tolerance to salinity and low temperature in transgenic tobacco producing glycine betaine. Journal of Experimental Botany, 51(343), Hooker, T. S., & Thorpe, T. A. (1998). Effects of fluridone and abscisic acid on lateral root initiation and root elongation of excised tomato roots cultured in vitro. Plant cell, tissue and organ culture, 52(3), Hsiao, T. C., & Xu, L. K. (2000). Sensitivity of growth of roots versus leaves to water stress: biophysical analysis and relation to water transport. Journal of Experimental Botany, 51(350), Irizarry, R. A., Wu, Z., & Jaffee, H. A. (2006). Comparison of Affymetrix GeneChip expression measures. Bioinformatics, 22(7), Ishida, T., Kurata, T., Okada, K., & Wada, T. (2008). A genetic regulatory network in the development of trichomes and root hairs. Annu. Rev. Plant Biol., 59,

159 Iyer-Pascuzzi, A. S., & Benfey, P. N. (2009). Transcriptional networks in root cell fate specification. Biochimica et Biophysica Acta (BBA)-Gene Regulatory Mechanisms, 1789(4), Jibran, R., Hunter, D. A., & Dijkwel, P. P. (2013). Hormonal regulation of leaf senescence through integration of developmental and stress signals. Plant molecular biology, 82 (6), Jithesh, M. N., Prashanth, S. R., Sivaprakash, K. R., & Parida, A. (2006). Monitoring expression profiles of antioxidant genes to salinity, iron, oxidative, light and hyperosmotic stresses in the highly salt tolerant grey mangrove, Avicennia marina (Forsk.) Vierh. by mrna analysis. Plant cell reports, 25(8), Johnston, D. S., & Nüsslein-Volhard, C. (1992). The origin of pattern and polarity in the Drosophila embryo. Cell, 68(2), Kagale, S., Divi, U. K., Krochko, J. E., Keller, W. A., & Krishna, P. (2007). Brassinosteroid confers tolerance in Arabidopsis thaliana and Brassica napus to a range of abiotic stresses. Planta, 225(2), Kaplan-Levy, R. N., Brewer, P. B., Quon, T., & Smyth, D. R. (2012). The trihelix family of transcription factors light, stress and development. Trends in plant science, 17(3), Keller, F., & Ludlow, M. M. (1993). Carbohydrate metabolism in drought-stressed leaves of pigeonpea (Cajanus cajan). Journal of Experimental Botany, 44(8), Khatun, S., & Flowers, T. J. (1995). Effects of salinity on seed set in rice. Plant, Cell & Environment, 18(1),

160 Khripach, V. A., Zhabinskii, V. N., & de Groot, A. E. (1998). Brassinosteroids: a new class of plant hormones. Academic Press. Kiegle, E., Moore, C. A., Haseloff, J., Tester, M. A., & Knight, M. R. (2000). Celltype specific calcium responses to drought, salt and cold in the Arabidopsis root. The Plant Journal, 23(2), Kim, S. G., Lee, A. K., Yoon, H. K., & Park, C. M. (2008). A membrane bound NAC transcription factor NTL8 regulates gibberellic acid mediated salt signaling in Arabidopsis seed germination. The Plant Journal, 55(1), Kuiper, D., Schuit, J., & Kuiper, P. J. C. (1990). Actual cytokinin concentrations in plant tissue as an indicator for salt resistance in cereals. Plant and Soil, 123(2), Labidi, N., Ammari, M., Mssedi, D., Benzerti, M., Snoussi, S., & Abdelly, C. (2010). Salt excretion in Suaeda fruticosa. Acta Biologica Hungarica, 61(3), Laux, T., Mayer, K. F., Berger, J., & Jurgens, G. (1996). The WUSCHEL gene is required for shoot and floral meristem integrity in Arabidopsis. Development, 122(1), Leek, J. T., Monsen, E., Dabney, A. R., & Storey, J. D. (2006). EDGE: extraction and analysis of differential gene expression. Bioinformatics, 22(4), Letham, D. S. (1963). Regulators of cell division in plant tissues: I. Inhibitors and stimulants of cell division in developing fruits: Their properties and activity in relation to the cell division period. New Zealand Journal of Botany, 1(3), Lilly, B., Galewsky, S., Firulli, A. B., Schulz, R. A., & Olson, E. N. (1994). D-MEF2: a MADS box transcription factor expressed in differentiating mesoderm and muscle cell 143

161 lineages during Drosophila embryogenesis. Proceedings of the National Academy of Sciences, 91(12), Lin, C. C., & Kao, C. H. (1995). NaCl stress in rice seedlings: starch mobilization and the influence of gibberellic acid on seedling growth. Botanical Bulletin of Academia Sinica, 36, Liu, J., & Zhu, J. K. (1998). A calcium sensor homolog required for plant salt tolerance. Science, 280(5371), Liu, J., Ishitani, M., Halfter, U., Kim, C. S., & Zhu, J. K. (2000). The Arabidopsis thaliana SOS2 gene encodes a protein kinase that is required for salt tolerance. Proceedings of the National Academy of Sciences, 97(7), Liu, Z. B., Ulmasov, T., Shi, X., Hagen, G., & Guilfoyle, T. J. (1994). Soybean GH3 promoter contains multiple auxin-inducible elements. The Plant Cell Online, 6(5), Lovén, J., Hoke, H. A., Lin, C. Y., Lau, A., Orlando, D. A., Vakoc, C. R.,... & Young, R. A. (2013). Selective inhibition of tumor oncogenes by disruption of superenhancers. Cell, 153(2), Lozano-Rodriguez, E., Hernandez, L. E., Bonay, P., & Carpena-Ruiz, R. O. (1997). Distribution of cadmium in shoot and root tissues1. Journal of Experimental Botany, 48(1), Luan, S., Lan, W., & Chul Lee, S. (2009). Potassium nutrition, sodium toxicity, and calcium signaling: connections through the CBL CIPK network. Current opinion in plant biology, 12(3),

162 Lutts, S., Kinet, J. M., & Bouharmont, J. (1995). Changes in plant response to NaCl during development of rice (Oryza sativa L.) varieties differing in salinity resistance. Journal of Experimental Botany, 46(12), Ma, S., & Bohnert, H. J. (2007). Integration of Arabidopsis thaliana stress-related transcript profiles, promoter structures, and cell-specific expression. Genome biology, 8(4), R49. Macler, B. A., & MacElroy, R. D. (1989). Productivity and food value of Amaranthus cruentus under non-lethal salt stress. Advances in Space Research, 9(8), Martınez-Antonio, A., & Collado-Vides, J. (2003). Identifying global regulators in transcriptional regulatory networks in bacteria. Current opinion in microbiology, 6(5), Maruyama, K., Todaka, D., Mizoi, J., Yoshida, T., Kidokoro, S., Matsukura, S.,... & Yamaguchi-Shinozaki, K. (2012). Identification of cis-acting promoter elements in coldand dehydration-induced transcriptional pathways in Arabidopsis, rice, and soybean. DNA research, 19(1), Mauromicale, G., & Licandro, P. (2002). Salinity and temperature effects on germination, emergence and seedling growth of globe artichoke. Agronomie-Sciences des Productions Vegetales et de l'environnement, 22(5), Merika, M., & Thanos, D. (2001). Enhanceosomes. Current opinion in genetics & development, 11(2), Michael, T. P., Breton, G., Hazen, S. P., Priest, H., Mockler, T. C., Kay, S. A., & Chory, J. (2008). A morning-specific phytohormone gene expression program underlying rhythmic plant growth. PLoS biology, 6(9), e

163 Munns, R. (2002). Comparative physiology of salt and water stress. Plant, Cell & Environment, 25(2), Munns, R., & Passioura, J. B. (1984). Hydraulic resistance of plants. III. Effects of NaCl in barley and lupin. Functional Plant Biology, 11(5), Narusaka, Y., Nakashima, K., Shinwari, Z. K., Sakuma, Y., Furihata, T., Abe, H.,... & Yamaguchi Shinozaki, K. (2003). Interaction between two cis acting elements, ABRE and DRE, in ABA dependent expression of Arabidopsis rd29a gene in response to dehydration and high salinity stresses. The Plant Journal, 34(2), Nemhauser, J. L., Hong, F., & Chory, J. (2006). Different plant hormones regulate similar processes through largely non-overlapping transcriptional responses. Cell, 126(3), Outlaw, Jr, W. H. (2003). Integration of cellular and physiological functions of guard cells. Critical Reviews in Plant Sciences, 22(6), Parida, A. K., & Das, A. B. (2005). Salt tolerance and salinity effects on plants: a review. Ecotoxicology and environmental safety, 60(3), Passioura, J. B., & Munns, R. (2000). Rapid environmental changes that affect leaf water status induce transient surges or pauses in leaf expansion rate. Functional Plant Biology, 27(10), Pathan, M. S., Subudhi, P. K., Courtois, B., & Nguyen, H. T. (2004). Molecular dissection of abiotic stress tolerance in sorghum and rice. Physiology and biotechnology integration for plant breeding. ebook ISBN: Plett, D., Johnson, A., Jacobs, A., & Tester, M. (2010). Cell type-specific expression of sodium transporters improves salinity tolerance of rice. GM crops, 1(5),

164 Pruneda-Paz, J. L., Breton, G., Para, A., & Kay, S. A. (2009). A functional genomics approach reveals CHE as a component of the Arabidopsis circadian clock. Science, 323(5920), Qasim, M., & Ashraf, M. (2006). Time course of ion accumulation and its relationship with the salt tolerance of two genetically diverse lines of canola (Brassica napus L.). Pakistan Journal of Botany, 38(3), 663. Rademacher, W. (2000). Growth retardants: effects on gibberellin biosynthesis and other metabolic pathways. Annual review of plant biology, 51(1), Reece-Hoyes, J. S., Diallo, A., Lajoie, B., Kent, A., Shrestha, S., Kadreppa, S., & Walhout, A. J. (2011). Enhanced yeast one-hybrid assays for high-throughput genecentered regulatory network mapping. Nature methods, 8(12), Rhoades, J. D., & Loveday, J. (1990). Salinity in irrigated agriculture. Agronomy, (30), Riechmann, J. L. (2002). Transcriptional regulation: a genomic overview. The Arabidopsis book, 1, e0085. Riechmann, J. L., Heard, J., Martin, G., Reuber, L., Keddie, J., Adam, L. & Yu, G. L. (2000). Arabidopsis transcription factors: genome-wide comparative analysis among eukaryotes. Science, 290(5499), Riechmann, L., & Winter, G. (2006). Early protein evolution: building domains from ligand-binding polypeptide segments. Journal of molecular biology, 363(2), Roppolo, D., De Rybel, B., Tendon, V. D., Pfister, A., Alassimone, J., Vermeer, J. E.,... & Geldner, N. (2011). A novel protein family mediates Casparian strip formation in the endodermis. Nature, 473(7347),

165 Rubio, F., Gassmann, W., & Schroeder, J. I. (1995). Sodium-driven potassium uptake by the plant potassium transporter HKT1 and mutations conferring salt tolerance. Science, 270(5242), Rubio, F., Gassmann, W., & Schroeder, J. I. (1995). Sodium-driven potassium uptake by the plant potassium transporter HKT1 and mutations conferring salt tolerance. Science, 270(5242), Saeed, A. I., Sharov, V., White, J., Li, J., Liang, W., Bhagabati, N.,... & Quackenbush, J. (2003). TM4: a free, open-source system for microarray data management and analysis. Biotechniques, 34 (2), 374. Sakakibara, H. (2006). Cytokinins: activity, biosynthesis, and translocation. Annu. Rev. Plant Biol., 57, Sato, T., & Theologis, A. (1989). Cloning the mrna encoding 1-aminocyclopropane-1- carboxylate synthase, the key enzyme for ethylene biosynthesis in plants. Proceedings of the National Academy of Sciences, 86(17), Savouré, A., Hua, X. J., Bertauche, N., Van Montagu, M., & Verbruggen, N. (1997). Abscisic acid-independent and abscisic acid-dependent regulation of proline biosynthesis following cold and osmotic stresses in Arabidopsis thaliana. Molecular and General Genetics MGG, 254(1), Schoof, H., Lenhard, M., Haecker, A., Mayer, K. F., Jürgens, G., & Laux, T. (2000). The Stem Cell Population of Arabidopsis Shoot Meristems Is Maintained by a Regulatory Loop between the CLAVATA and WUSCHEL Genes. Cell, 100(6), Schultz, E. A., & Haughn, G. W. (1991). LEAFY, a homeotic gene that regulates inflorescence development in Arabidopsis. The Plant Cell Online, 3(8),

166 Schützendübel, A., & Polle, A. (2002). Plant responses to abiotic stresses: heavy metal induced oxidative stress and protection by mycorrhization. Journal of experimental botany, 53 (372), Schwechheimer, C. (2008). Understanding gibberellic acid signaling are we there yet?. Current opinion in plant biology, 11(1), Scott, M. P. (2000). Development: the natural history of genes. Cell, 100(1), Shabala, S., & Cuin, T. A. (2008). Potassium transport and plant salt tolerance.physiologia Plantarum, 133(4), Shabala, S., Cuin, T. A., & Pottosin, I. (2007). Polyamines prevent NaCl-induced K + efflux from pea mesophyll by blocking non-selective cation channels. FEBS letters, 581(10), Shahbaz, M., & Ashraf, M. (2008). Does exogenous application of 24-epibrassinolide ameliorate salt induced growth inhibition in wheat (Triticum aestivum L.)? Plant Growth Regulation, 55(1), Shannon, P., Markiel, A., Ozier, O., Baliga, N. S., Wang, J. T., Ramage, D.,... & Ideker, T. (2003). Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome research, 13(11), Shen, Q., & Ho, T. H. (1995). Functional dissection of an abscisic acid (ABA)-inducible gene reveals two independent ABA-responsive complexes each containing a G-box and a novel cis-acting element. The Plant Cell Online, 7(3), Shi, H., Ishitani, M., Kim, C., & Zhu, J. K. (2000). The Arabidopsis thaliana salt tolerance gene SOS1 encodes a putative Na + /H + antiporter. Proceedings of the national academy of sciences, 97(12),

167 Shi, H., Kim, Y., Guo, Y., Stevenson, B., & Zhu, J. K. (2003). The Arabidopsis SOS5 locus encodes a putative cell surface adhesion protein and is required for normal cell expansion. The Plant Cell Online, 15(1), Shi, H., Xiong, L., Stevenson, B., Lu, T., & Zhu, J. K. (2002). The Arabidopsis salt overly sensitive 4 mutants uncover a critical role for vitamin B6 in plant salt tolerance. The Plant Cell Online, 14(3), Shiu, S. H., Shih, M. C., & Li, W. H. (2005). Transcription factor families have much higher expansion rates in plants than in animals. Plant Physiology, 139(1), Siriwardana, N. S., & Lamb, R. S. (2012). The poetry of reproduction: the role of LEAFY in Arabidopsis thaliana flower formation. Int. J. Dev. Biol, 56, Skirycz, A., Claeys, H., De Bodt, S., Oikawa, A., Shinoda, S., Andriankaja, M.,... & Inzé, D. (2011). Pause-and-stop: the effects of osmotic stress on cell proliferation during early leaf development in Arabidopsis and a role for ethylene signaling in cell cycle arrest. The Plant Cell Online, 23(5), Smyth, G. K. (2005). Limma: linear models for microarray data. In Bioinformatics and computational biology solutions using R and Bioconductor (pp ). Springer New York. Soltis, D. E., Chanderbali, A. S., Kim, S., Buzgo, M., & Soltis, P. S. (2007). The ABC model and its applicability to basal angiosperms. Annals of Botany, 100(2), Spychalla, J. P., & Desborough, S. L. (1990). Superoxide dismutase, catalase, and α- tocopherol content of stored potato tubers. Plant Physiology, 94(3),

168 Stepanova, A. N., Robertson-Hoyt, J., Yun, J., Benavente, L. M., Xie, D. Y., Doležal, K.,... & Alonso, J. M. (2008). TAA1-Mediated Auxin Biosynthesis Is Essential for Hormone Crosstalk and Plant Development. Cell, 133(1), Stewart, G. R., & Lee, J. A. (1974). The role of proline accumulation in halophytes. Planta, 120(3), Storey, J. D., & Tibshirani, R. (2003). Statistical significance for genomewide studies. Proceedings of the National Academy of Sciences, 100(16), Storey, J. D., Xiao, W., Leek, J. T., Tompkins, R. G., & Davis, R. W. (2005). Significance analysis of time course microarray experiments. Proceedings of the National Academy of Sciences of the United States of America, 102(36), Struhl, K. (1999). Fundamentally different logic of gene regulation in eukaryotes and prokaryotes. Cell, 98(1), 1-4. Sugimoto, M., & Sakamoto, W. (1997). Putative phospholipid hydroperoxide glutathione peroxidase gene from Arabidopsis thaliana induced by oxidative stress. Genes & genetic systems, 72(5), Sun, K., & Hauser, B. (2001). Salt stress induces anatomical changes in ovules and embryos, ultimately resulting in seed abortion. In The 12th International Meeting on Arabidopsis Research. June (pp ). Sun, T. P. (2008). Gibberellin metabolism, perception and signaling pathways in Arabidopsis. The Arabidopsis Book, 6, e0103. Tran, L. S. P., Urao, T., Qin, F., Maruyama, K., Kakimoto, T., Shinozaki, K., & Yamaguchi-Shinozaki, K. (2007). Functional analysis of AHK1/ATHK1 and cytokinin 151

169 receptor histidine kinases in response to abscisic acid, drought, and salt stress in Arabidopsis. Proceedings of the National Academy of Sciences, 104(51), Trejo, C. L., & Davies, W. J. (1991). Drought-induced closure of Phaseolus vulgaris L. stomata precedes leaf water deficit and any increase in xylem ABA concentration. Journal of Experimental Botany, 42(12), Ueguchi-Tanaka, M., Ashikari, M., Nakajima, M., Itoh, H., Katoh, E., Kobayashi, M.,... & Matsuoka, M. (2005). GIBBERELLIN INSENSITIVE DWARF1 encodes a soluble receptor for gibberellin. Nature, 437(7059), Ulmasov, T., Murfett, J., Hagen, G., & Guilfoyle, T. J. (1997). Aux/IAA proteins repress expression of reporter genes containing natural and highly active synthetic auxin response elements. The Plant Cell Online, 9(11), van der Graaff, E., Laux, T., & Rensing, S. A. (2009). The WUS homeobox-containing (WOX) protein family. Genome Biol, 10 (12), 248. Vyroubalová, Š., Václavíková, K., Turečková, V., Novák, O., Šmehilová, M., Hluska, T.,... & Galuszka, P. (2009). Characterization of new maize genes putatively involved in cytokinin metabolism and their expression during osmotic stress in relation to cytokinin levels. Plant Physiology, 151(1), Wang, Y., Li, L., Ye, T., Zhao, S., Liu, Z., Feng, Y. Q., & Wu, Y. (2011). Cytokinin antagonizes ABA suppression to seed germination of Arabidopsis by downregulating ABI5 expression. The Plant Journal, 68(2), Wang, Y., Liu, C., Li, K., Sun, F., Hu, H., Li, X.,... & Liu, M. (2007). Arabidopsis EIN2 modulates stress response through abscisic acid response pathway. Plant molecular biology, 64(6),

170 Wang, Z. Y. (2012). Brassinosteroids modulate plant immunity at multiple levels. Proceedings of the National Academy of Sciences, 109(1), 7-8. Weigel, D., & Meyerowitz, E. M. (1994). The ABCs of floral homeotic genes. Cell, 78(2), Wen, C. K., & Chang, C. (2002). Arabidopsis RGL1 encodes a negative regulator of gibberellin responses. The Plant Cell Online, 14(1), Werner, T., Motyka, V., Laucou, V., Smets, R., Van Onckelen, H., & Schmülling, T. (2003). Cytokinin-deficient transgenic Arabidopsis plants show multiple developmental alterations indicating opposite functions of cytokinins in the regulation of shoot and root meristem activity. The Plant Cell Online, 15(11), White, P. J. (1999). The molecular mechanism of sodium influx to root cells. Trends in plant science, 4(7), Whyte, W. A., Orlando, D. A., Hnisz, D., Abraham, B. J., Lin, C. Y., Kagey, M. H.,... & Young, R. A. (2013). Master transcription factors and mediator establish superenhancers at key cell identity genes. Cell, 153(2), Willekens, H., Chamnongpol, S., Davey, M., Schraudner, M., Langebartels, C., Van Montagu, M.,... & Van Camp, W. (1997). Catalase is a sink for H2O2 and is indispensable for stress defence in C3 plants. The EMBO journal, 16(16), Wu, S. J., Ding, L., & Zhu, J. K. (1996). SOS1, a genetic locus essential for salt tolerance and potassium acquisition. The Plant Cell Online, 8(4), Xiong, L., Schumaker, K. S., & Zhu, J. K. (2002). Cell signaling during cold, drought, and salt stress. The Plant Cell Online, 14(suppl 1), S165-S

171 Xu, W., Jia, L., Shi, W., Liang, J., Zhou, F., Li, Q., & Zhang, J. (2013). Abscisic acid accumulation modulates auxin transport in the root tip to enhance proton secretion for maintaining root growth under moderate water stress. New Phytologist, 197(1), Yamaguchi-Shinozaki, K., & Shinozaki, K. (2005). Organization of cis-acting regulatory elements in osmotic-and cold-stress-responsive promoters. Trends in plant science, 10(2), Yokoi, S., Quintero, F. J., Cubero, B., Ruiz, M. T., Bressan, R. A., Hasegawa, P. M., & Pardo, J. M. (2002). Differential expression and function of Arabidopsis thaliana NHX Na + /H + antiporters in the salt stress response. The Plant Journal, 30(5), Zeevaart, J. A. D., & Creelman, R. A. (1988). Metabolism and physiology of abscisic acid. Annual review of plant physiology and plant molecular biology, 39(1), Zhang, H., Han, W., De Smet, I., Talboys, P., Loya, R., Hassan, A.,... & Wang, M. H. (2010). ABA promotes quiescence of the quiescent centre and suppresses stem cell differentiation in the Arabidopsis primary root meristem. The Plant Journal, 64(5), Zhao, X. C., & Schaller, G. E. (2004). Effect of salt and osmotic stress upon expression of the ethylene receptor ETR1 in Arabidopsis thaliana. FEBS letters, 562(1), Zhong, R., Lee, C., & Ye, Z. H. (2010). Evolutionary conservation of the transcriptional network regulating secondary cell wall biosynthesis. Trends in plant science, 15(11), Zhu, J. K. (2000). Genetic analysis of plant salt tolerance using Arabidopsis. Plant Physiology, 124(3), Zhu, J. K. (2001). Plant salt tolerance. Trends in plant science, 6(2),

172 Zhu, J. K., Liu, J., & Xiong, L. (1998). Genetic analysis of salt tolerance in Arabidopsis: evidence for a critical role of potassium nutrition. The Plant Cell Online, 10(7), Zinzen, R. P., Girardot, C., Gagneur, J., Braun, M., & Furlong, E. E. (2009). Combinatorial binding predicts spatio-temporal cis-regulatory activity. Nature, 462(7269), Zou, C., Sun, K., Mackaluso, J. D., Seddon, A. E., Jin, R., Thomashow, M. F., & Shiu, S. H. (2011). Cis-regulatory code of stress-responsive transcription in Arabidopsis thaliana. Proceedings of the National Academy of Sciences, 108(36),

173 Appendix Appendix Figure. Summary of the patterns of tissue-specific cis-regulatory elements. Bar charts show the expression levels of the two genes that I used to generate the synthetic promoters, and it was generated based on the transcriptional root map (Brady et al., 2007). 156